1/1/2025 - Certification for AI Readiness Ensuring Preparedness in Business

Description

Chapters

Transcript

0:00 good
0:01 people good people of the internet
0:21 [Music]
0:35 [Music]
0:37 good day good day good day good
0:42 [Music]
0:57 day there's been something baby i' try
1:00 to
1:04 say aging it seems I don't know
1:10 how past in the future now surrounding
1:16 me surrender to whatever treat thr can
1:20 be
1:21 [Music]
1:22 found there's been little
1:26 trouble since you came to my rescue
1:30 [Music]
1:35 if you of the rest I you long ago but I
1:41 couldn't do
1:43 [Music]
1:45 that oh tell me now
1:49 wom never went to
1:53 well make a man crazy make him cold as
1:57 hell
2:01 I know woman you wish me
2:05 well but
2:07 try still going to have to find my own
2:12 way
2:16 through happy New Year good people happy
2:20 New Year to all of you on the internet
2:24 who love AI or hate AI or afraid of AI
2:28 or don't even know what AI is
2:31 [Music]
2:34 my name's Kyle Shannon this is the AI
2:36 learning lab we've got singing dogs
2:40 strumming
2:41 men and pink
2:45 bows we're not afraid of a pink bow the
2:47 AI learning
2:51 lab I watch football all day
2:58 [Music]
3:10 so if you have questions about AI if you
3:12 have thoughts about AI oh the title of
3:16 my Tik Tok video says 2024 oh the title
3:19 of my streamyard video says 2024 it's
3:24 2025 how long are we gonna [ __ ] make
3:27 that mistake
3:32 [Music]
3:34 today is the first day in at least a
3:36 month of writing the wrong date on
3:39 [Music]
3:45 things um if you're if you're followers
3:48 of David Shapiro he put out a pile of
3:51 posts today on X that I actually thought
3:54 I reposted a bunch of
3:57 them um I thought they're actually quite
3:59 important
4:00 [Music]
4:02 and and there was a lot of similar kind
4:06 of
4:07 sentiment that
4:11 um that that the
4:14 acceleration is really about to
4:16 start and what the acceleration is from
4:19 is when these tools get good enough that
4:22 they can start improving themselves or
4:24 they can start taking action on their
4:26 own and rather than having to hire a
4:29 bunch of people people to go solve some
4:31 problems you can just turn a bunch of AI
4:33 agents loose to go solve them then you
4:36 start solving problems faster and faster
4:39 and the the the quality of the
4:41 technology goes up and the the speed of
4:44 those improvements increases as well so
4:48 it seems pretty clear that that's going
4:50 to happen
4:53 [Music]
5:01 [Music]
5:10 what was interesting about his posts is
5:12 they
5:13 were one of them was was was directly
5:17 about getting a divorce from work
5:21 life because we're not going to have to
5:24 now I don't think that's going to happen
5:26 in 2025 I don't think that's going to
5:29 happen in 2026 because I think
5:31 businesses and people move slower than
5:34 technology
5:36 but I think it's important for us to be
5:39 ready for it I think it's important for
5:41 us to be
5:43 ready
5:45 that that the AI is not going to
5:49 necessarily require us to
5:52 participate and and that will start to
5:54 experience in 2025 and that's that's
5:58 agents um
6:01 so I think any of the
6:04 existential questions that we've had
6:06 about AI but where we still have to
6:09 initiate the
6:12 contact I think I think the stakes go up
6:15 in
6:16 [Music]
6:24 2025 Lord digital Gods thank you for the
6:27 gifts the Roses Are Much appreciated
6:30 have you tried using do browser yet AI
6:35 agent Chrome extension I haven't Mike um
6:39 I have not um I am I tend to not be a
6:42 fan of Chrome extensions not because I
6:45 particularly give a [ __ ] about Chrome
6:47 extensions one way or the other but as
6:49 an ADHD person what I tend to do is
6:53 install Chrome extensions and then
6:55 forget about them and then 17 years
6:59 later I've got 270 Chrome extensions and
7:02 I can't figure out why chrome doesn't
7:04 [Laughter]
7:07 [Music]
7:16 work do browser got it how can we get on
7:19 the 11: a call tomorrow good question I
7:23 I know I put out a a video saying
7:25 there's going to be an 11: a. call
7:27 tomorrow and I didn't say [ __ ] about
7:29 where to find it
7:30 uh partially because we're flying This
7:32 Plane as we're building it as we talked
7:34 about in AI festivas
7:36 um we're we're going up on restream uh
7:40 an Murphy's restream so I will do one of
7:42 these Tik Tock things so you can come to
7:44 the Tik Tok Channel you can also come to
7:47 LinkedIn and it will be on LinkedIn I'll
7:49 make sure that I I post it on my
7:52 LinkedIn in fact if you go to linked in
7:54 right now you can actually register for
7:56 the event and I think that will get you
7:59 an email with
8:01 the with the details so and it'll
8:05 probably be on YouTube Somewhere YouTube
8:07 LinkedIn and Tick Tock the AI learning
8:10 lab on Tik
8:11 [Music]
8:23 Tock there's something really
8:25 interesting that happened largely as a
8:27 result of AI FES of us um I've had a bit
8:30 of a breakthrough I've I've been
8:32 struggling for the past six months or
8:35 so
8:37 um figuring out my role the role of AI
8:40 learning lab the role of AI Salon like
8:42 like what is the like what do people
8:45 need and I feel like what they needed in
8:48 the in the first year was here's what
8:52 these tools are some version of that
8:56 right and and while people that are new
8:59 to AI are going to need that I feel like
9:01 there's there's plenty of resources out
9:03 there now that are about that and so I
9:06 think I think you know it's not a it's
9:09 not a desert of knowledge on that
9:12 [Music]
9:17 front and one of the things we talked
9:20 about in the AI
9:23 Salon one of the things that's kind of
9:25 the
9:26 core of the AI salon
9:30 is what we were calling the journey to
9:32 AI
9:35 literacy Tik Tock
9:38 pin um do you use the premium version of
9:41 simple AI I do not I I just downloaded
9:43 it once I I downloaded it used it once I
9:47 almost used it
9:49 today and then I chickened
9:54 out I I I still have
9:57 like weirdness about so what simple AI
10:00 is if you don't know it's an app that
10:03 you get and it kind of has a chat GPT
10:06 like interface and then you can say hey
10:08 call up you know all the businesses near
10:11 me that sell hardware and see if they
10:13 have a plumbing snake or whatever and it
10:17 will make phone calls it'll it'll make
10:18 all these phone calls simultaneously and
10:21 record the transcript of the
10:22 conversation with the person that picked
10:24 up the phone and then give you the
10:26 answer like which store has it um it's
10:29 pretty remarkable but no I do not have
10:31 the paid version of it because I just I
10:33 have I have not quite come to the I have
10:37 not quite
10:39 reconciled letting robots make calls on
10:41 my behalf yet I will I'll get there it's
10:45 just still weird to
10:46 me
10:48 anyway this so in the AI Salon we we
10:51 always talked about this thing the
10:53 journey to AI literacy and it was play
10:55 first mindfully create generously lead
10:58 that part hasn't changed the the the
11:00 three steps but what has changed is is
11:03 the name of the of the process the
11:06 journey journey to AI
11:10 literacy a remarkable thing that
11:13 happened during AI festiv is remarkable
11:16 to me was that every single session
11:20 every single session even if they were
11:23 dealing with
11:25 technical
11:28 explanations was more about the person
11:31 and what they were trying to do in the
11:32 world than it was about the
11:34 technology like even P modder that did
11:37 this thing on how llms work he was he he
11:41 did it in this Spirit of listen I'm like
11:44 way up here in terms of you know
11:48 mathematical knowledge you know he's
11:50 he's a doctor he's a PhD of of
11:52 mathematics that works at Google and he
11:55 created this presentation that like
11:56 brought it all the way down and he said
11:58 you know I think what people need right
11:59 right now and and every to a person
12:03 every single person
12:05 was some version of I'm trying to figure
12:09 this stuff out like here's what I do
12:12 here's what I know and here's how I'm
12:14 using AI not here's how to use this tool
12:18 right and what and an an Murphy brought
12:22 this up she said she said there's
12:24 something about AI literacy that is too
12:27 literal right that it's it's too much
12:29 about the tools and so what we ended up
12:31 we went back and forth and had I don't
12:33 know a dozen different names and what we
12:36 landed on was this let me share my
12:38 screen for a second here and I'm sorry
12:40 my hair looks like a carpet salesman but
12:42 you know what are you gonna
12:44 do
12:45 um journey to AI
12:48 Readiness and so the
12:52 um
12:57 the um the certification that we're
12:59 doing based on um AI um AI Festivus is
13:05 is going to be an AI Readiness
13:08 certification here's why I like this
13:11 term instead of AI literacy I actually
13:14 now do not believe that it's possible to
13:18 become AI literate it's moving too fast
13:21 there's too many
13:22 tools like even if you just wanted to
13:25 learn chat GPT like I like it's it is it
13:29 would be a struggle to really learn all
13:31 of what cat GPT can
13:34 do same with Claude same with all of
13:36 them like all of the stuff that Google
13:39 has out there they've got so much stuff
13:41 out there it's just this ecosystem of of
13:46 functionality and
13:47 so and then even if you went and learned
13:50 all that you're not learning all the
13:53 other [ __ ] that's happening and then at
13:56 some point Google's going to release
13:57 something that trumps all of the [ __ ]
13:59 that you learned and now you're going to
14:00 have to relearn it right so the idea of
14:03 AI literacy isn't there but being ready
14:05 for what's coming from AI that I think
14:07 we can train to and March towards and so
14:11 what this certification is going to be
14:13 about is re is really that it's it's
14:15 more mindset Cindy [ __ ] calls it the AI
14:18 mindset you know are you in an AI
14:20 mindset are you curious are you
14:22 adventurous are you
14:24 adaptable are you willing to you know
14:27 slay the sacred cows of of whatever of
14:29 your own personal selfworth of the
14:32 definition of work of the definition of
14:34 a
14:36 company being AI ready means that you're
14:43 in I mean you know I we use the metaphor
14:46 a lot there's like these big waves
14:48 coming and it's better to be on a
14:51 surfboard than on the beach and that
14:54 that kind of feels the most right that
14:56 you got to figure out what surfing is
14:58 and what it means to be on a surfboard
15:00 and what it means to stand up on one and
15:03 what it means to stand up on one when
15:04 you're in a wave and what it means to
15:06 fall off one and what it means to get
15:08 back on even if you get the wind knocked
15:10 out of you at the bottom of the
15:14 ocean so journey to AI Readiness so so I
15:17 changed the the AI
15:19 website all
15:23 right so I don't know I don't know if
15:26 that's
15:28 valuable but happy New Year Happy New
15:30 Year to
15:35 everyone it's too fast to be
15:38 [Music]
15:46 literate oh I see Vicky just put if
15:48 you're on the YouTube channel in the
15:50 comments Vicky just put a link to the
15:53 LinkedIn post that's got the the sign up
15:56 for tomorrow's so tomorrow so I haven't
15:58 said it someone asked about it but I
15:59 haven't actually said it tomorrow at
16:01 11:00 a.m. mountain time 10:00 a.m.
16:03 Pacific Ann Murphy and I are going to do
16:05 an AI festiv us wrapup so we're gonna
16:08 say here's what we learned here's what
16:10 happened you know here's all the
16:13 insights here's what people have said
16:15 you can't believe the amount of notes
16:18 that we're getting from people who are
16:20 like this has already changed my life
16:23 like I had given up you know I was doing
16:26 AI for a while then I gave up and I just
16:28 got lost and and I'm back in and I have
16:30 clarity
16:32 now it's so cool it's so
16:35 cool so tomorrow at 11 you it's going to
16:38 be on LinkedIn I'm gonna I'm going to
16:40 fire up Tik Tock
16:43 live AI learning
16:46 lab and and I'll do like this where I'm
16:48 filming the screen so you can get all
16:50 those Mo a patterns and hurt your eyes
16:52 you're
16:54 welcome let me see if I can get those a
16:56 little less shitty
17:00 no no you're just G to have to deal if I
17:03 go closer does that
17:05 help that's better there you go you're
17:11 welcome all right yeah that's
17:14 [Music]
17:21 good Daisy May Clarity is what I came
17:24 away with I came away with Clarity and I
17:27 came away with hope
17:30 and I came away with Humanity
17:34 like the 34 people that presented to a
17:37 person were really [ __ ] impressive
17:40 people right like smart accomplished
17:45 like no [ __ ] that I don't have a [ __ ]
17:48 clue
17:50 about and to a
17:52 person it didn't feel
17:56 like anyone had it figured out
18:02 they were all in this inquiry they were
18:04 all in this I don't quite know what
18:06 surfing is but I figured up how to get
18:09 up on a board on this little wave I'm
18:12 thinking about the big waves over there
18:13 but the the little one I sort of got for
18:16 now you know it had a Humanity to
18:20 it that was it wasn't a surprise because
18:23 like one of the things about the AI
18:26 Salon community and the she she leads AI
18:28 community
18:30 um and some of the other communities we
18:31 had speakers come in from is
18:36 that they tend to be full of empathetic
18:40 generous interesting smart people so it
18:44 doesn't surprise me that you know the
18:46 speakers that emerged out of those
18:49 communities have those
18:51 traits but I don't I think I was
18:55 expecting
18:57 more AI literacy
19:00 talk I figured it out here's how you do
19:02 it here's the templates here's the
19:03 prompting strategy here's the this
19:05 here's the that right even the people
19:08 that had prompting strategies in their
19:10 conversation it wasn't about the
19:12 prompting strategy it was about what
19:14 they were trying to accomplish as a
19:16 human being it was about their creative
19:19 imperative
19:21 right really really cool so that's
19:25 tomorrow 10:00 a.m. Pacific 11:00 a.m.
19:28 here and then we're going to talk about
19:30 so one of the requests that that
19:32 happened multiple times throughout the
19:34 day um on AI festiv us was people were
19:37 saying is this a certification or can
19:41 this be a certification or will this be
19:42 a certification or can we get certified
19:45 on this stuff
19:47 and there was enough of it that it that
19:50 and and by the end I'm like there was so
19:56 much so much was communic at that even
20:01 if you sat there for the full 36 hours
20:04 most of it's going to wash over your
20:05 head so I think doing an intentional
20:08 Deep dive on all the sessions or all the
20:11 sessions in a track I think there was a
20:13 clear creative track there was a clear
20:15 business track and then there were some
20:17 some miscellaneous maybe just sort of
20:18 General AI stuff like you know policy
20:21 and security and and you know how llms
20:24 work and things like that so I think
20:26 there was kind of a general track and so
20:29 I so what we're planning on right now is
20:32 breaking the sessions up into tracks and
20:35 then you can subscribe to the entire if
20:38 you want to go do all of it or if you
20:39 want to just do it track by track you
20:41 can do
20:43 that uh SE slug of Doom I was definitely
20:45 looking for that technical strategy yeah
20:50 yeah there wasn't there wasn't this this
20:53 was not this was not
20:57 that and if you were looking for that
20:59 se- slug then that that might have been
21:02 a let down but here's the thing about
21:07 that I mean I would strongly encourage
21:09 you to go read David Shapiro's posts
21:11 from
21:12 today um on Twitter because a lot of
21:15 what he talked about
21:18 is he feels that he no longer has
21:21 anything to contribute
21:23 technically like this is a guy
21:27 who was the guy to go to when you need
21:29 to deal with something something
21:33 technically and he's like with agents
21:35 and reasoning basically what's coming in
21:38 in the next three months is going to be
21:40 the beginnings of it
21:44 [Music]
21:48 um a lot of getting that stuff done is
21:51 just going to be needing to have the
21:53 intention to get it done not necessarily
21:56 needing to know how to set it up and I I
21:59 know that freaks a lot of people out
22:02 but given that assuming that that's true
22:05 and whether it's three months from now
22:07 or a year from now or three years from
22:10 now how do you how do you get ready for
22:12 that right how do you get ready for that
22:14 and I think that's what FES of us was
22:22 about so one thing I thought I would do
22:24 tonight um
22:28 Monday I tried to set up a notebook LM
22:31 with all the transcripts from the from
22:33 Festivus and
22:42 um because the videos were set to
22:45 private which they are they're only
22:47 available in the AI salon right now you
22:50 can't pull transcripts from them
22:52 automatically so um I ended up getting
22:56 the transcripts today and I so so I'm
22:58 going to build build a notebook LM
23:00 tonight with the
23:03 transcripts oh man Joy pie what's
23:05 happening good to see who we got in the
23:07 house we got Tobias we got Vicky Lori
23:11 Blair Cosmic
23:13 lovers
23:17 ripy Vicky Baptist Dr
23:20 J forever
23:23 hooked we
23:25 got fair amount of people on X hello X
23:29 what's happening hope you're doing well
23:31 Happy New Year to
23:34 everybody um so I figured we'll do a
23:36 little little hands on building some
23:38 [ __ ] and getting some insights from from
23:41 the
23:43 days I think it's actually quite
23:45 remarkable
23:47 that I you know what's funny is because
23:51 I've not been in Big Data
23:55 [Music]
23:57 um if you've been in big data and big
23:59 data analytics it is probably not super
24:02 remarkable to you that computers can
24:04 process a lot of data
24:06 [Laughter]
24:08 quickly but as just a regular human
24:11 being who's just been out here using
24:13 computers and being the subject of big
24:15 data right I I contribute to Google's
24:18 data pool and Facebook's data pool and
24:20 you know YouTube's data
24:23 pool um I've been a contributor to
24:26 rather than a user of the Big Data
24:29 [Music]
24:31 it still kind of melts my face that I
24:34 can upload transcripts of 24 hours of
24:38 video and it's just instantly available
24:41 it understands everything that happened
24:44 instantly we're going to be able to ask
24:46 it about any one of the
24:48 sessions any nuances any comments of any
24:52 kind we're going to be able to make
24:54 podcasts about day one day
24:57 two the whole thing
24:59 Kyle what is notebook LM well let's go
25:04 [Music]
25:07 play so if you want to play along you
25:09 can go to notebook lm.
25:13 google.com and it's it's not dissimilar
25:17 to if you've ever created a custom
25:20 GPT where you can upload data to a
25:23 custom GPT the difference here though is
25:26 that notebook LM is essentially limits
25:30 what it processes to the uploaded
25:33 knowledge it uses large language models
25:36 to be able to do lots of things with
25:38 that knowledge but like if if if I ask
25:41 it about you know some marketing AI
25:44 Institute event that Paul rater did the
25:47 same time that we did AI Salon if that
25:49 happened I don't know if it did it
25:50 wouldn't do it it would just say well
25:52 I've I've only got access to you know
25:54 the the data here so so it's it's kind
25:57 of a a really interesting way to
26:00 limit um a data set that you interact
26:03 with and in a in a weird way it's kind
26:05 of a I don't know if you've heard the
26:07 term rag but it rag stands for retrieval
26:10 augmented generation where where you can
26:13 take you can upload files effectively to
26:16 a database where they get embedded into
26:18 this database and then you can search on
26:21 those files and you can retrieve from
26:23 those files specifically it's kind of a
26:26 a poor man's version of that where where
26:29 you just upload a bunch of data and then
26:30 you can start interacting with it um it
26:32 still hallucinates it's not perfect um
26:35 but it's also got this really cool
26:37 feature called audio
26:39 overview so you upload all this data so
26:42 I'm going to upload all the transcripts
26:44 of all the days of you know the both
26:47 days of of the AI Fest of us um and and
26:51 we can make a podcast about the 24
26:55 sessions or the 23
26:57 sessions that's pretty slick right and
27:00 it sounds like um the podcast sounds
27:02 like you know an NPR style thing where
27:05 it's like the dude going well there's
27:06 this AI Festivus thing oh AI Festivus
27:09 that's really interesting they go back
27:11 and forth pretty slight it Dums down
27:13 white papers yeah it's it's the uh the
27:17 audio overview thing is shockingly
27:19 powerful
27:25 [Music]
27:30 Incognito I went to AI Festivus and all
27:32 I got was this lousy t-shirt we should
27:34 make an AI Festivus lousy
27:38 [Music]
27:43 t-shirt but anyway we're making the the
27:47 certification the the sort of AI
27:50 Festivus special edition um of of AI
27:54 Readiness certification and then I think
27:56 what what I'm excited about what what I
27:58 think do this idea of
28:01 of AI Readiness um as something you can
28:05 teach I'm really intrigued by it and I'm
28:08 I'm also intrigued to do it as the AI
28:10 salon and and as she leads AI are two
28:12 communities coming together and saying
28:14 you know here's what we think it takes
28:16 to be to be ready for what's coming to
28:19 the extent any of us can be ready for
28:21 what's
28:26 coming all right let's go to no book LM
28:29 shall we let's let's get let's get
28:30 geeking
28:34 [Music]
28:37 out so when you land on Notebook LM they
28:40 got some new uh some new uh interface
28:44 here which is kind of slick and nice you
28:46 can view it in tile mode or and I assume
28:49 I could put pictures in here if I wanted
28:50 it to look
28:52 better and you'll notice AI Festivus has
28:55 zero sources so we're going to start one
28:56 completely from scratch
28:59 and then you end up they they just
29:01 changed this interface you can see all
29:03 this right yeah am I good am I good tab
29:06 wise is everything good I don't have any
29:09 nasty notes from Brandon Vicky's not
29:11 yelling at
29:12 me my my mods over and Tik Tok aren't
29:15 yelling at me wait a minute something's
29:17 not right why am I not being yelled at
29:20 you can
29:22 make oh because I'm still I'm still
29:25 decently on track
29:28 [Music]
29:37 yeah 11:00 a.m. mountain time
29:40 [Music]
29:49 tomorrow Brandon is trying to force a
29:51 New Year's resolution down my throat
29:54 that I'm going to I'm going to follow
29:55 agendas in 2025 no I'm not
29:58 [Music]
29:58 [Laughter]
30:02 [Music]
30:11 side hustle Mimi I'm so not sober the
30:15 salon.
30:22 a daisy May new year new Kyle absolutely
30:26 not same old [ __ ]
30:31 [Music]
30:43 what's the drink tonight the drink
30:44 tonight is
30:46 emergency super orange or Tangerine
30:49 flavor in ice Rocky Mountain ice
30:56 water I've got this terrible drink
31:01 problem all right here's what we're
31:03 going to
31:06 [Music]
31:07 do
31:09 okay so the way this works is you add a
31:12 source so you got sources to the left
31:14 you have your basic chat window in the
31:16 middle and then you've got [ __ ] you
31:18 create on the right I think is their
31:21 General the general thing now now one
31:24 just weird AI uh ux thing that they did
31:28 is is because I have no sources they add
31:30 add a source to the middle panel but
31:32 that's not the sources panel the sources
31:34 panel is the left end but whatever let's
31:36 not get too picky shall we okay
31:40 um so normally you'd be able to put in a
31:43 YouTube url if the YouTube If the
31:45 YouTube is public so so here's a just a
31:48 cool thing if you didn't know that that
31:50 it can do
31:51 this you can go to YouTube and you can
31:55 go find a [ __ ] ton of YouTube videos on
31:58 a subject that you want to learn
31:59 something about just grab all those URLs
32:02 and come over here to notebook LM paste
32:05 the URLs in it will go out grab all the
32:07 transcripts from those videos bring them
32:10 back and then you can start asking it
32:12 questions about all those different
32:14 videos it's pretty slick but but because
32:18 of just technical reasons we can't do
32:20 that so what we're going to do is we're
32:21 going to
32:23 upload um some some source
32:26 files let's see
32:30 [Music]
32:36 [Music]
32:39 where's oh okay that can't be run as a
32:42 hang on a
32:43 sec I saved something as a word doc it
32:46 Google being Google oh that it's it's so
32:49 nice to know that corporate
32:51 douchebaggery is still in play it won't
32:54 open a do a word file I have to convert
32:57 it from word word to something
33:07 else oh my God okay you uh I'll just do
33:12 it as a text
33:14 file
33:15 oh y'all
33:19 y'all I guess I can do it as a PDF all
33:22 right I'll do it as a
33:26 PDF okay
33:30 oh corporate douchebaggery where far art
33:34 thou why so
33:36 douchy could they could they support a
33:39 Word document yes yes the word file
33:43 format has been around for
33:47 decades everybody knows how to read
33:51 them even Google all right so the first
33:54 thing I did was I uploaded the Run of
33:57 show wait what why did it go red oh oh
34:00 because it's PDF
34:02 okay so this run-of show document why I
34:06 uploaded that first is that's basically
34:09 the the schedule of the day so I don't
34:13 think that notebook LM cares order of
34:15 operations like I could probably upload
34:17 all these things in any order and it
34:19 wouldn't matter but I just have
34:24 this
34:26 suspicion that it matters
34:28 but it may not okay
34:31 now we're going to add some more
34:35 sources so I got a bunch of text files
34:38 let's sort them by
34:43 name and so they're all 1.0 1.2 1.3 1.1
34:48 was missing whatever it's
34:50 fine
34:52 um just a numbering
34:55 thing wow day two was more packed than
34:58 one I didn't realize that that's
35:00 cool all right so
35:05 nowo all those things are getting
35:11 inserted oh you're recovering 1.1 okay
35:14 so there might be one thing missing here
35:15 okay cool got
35:20 it something got
35:22 lost all right so now apparently
35:29 so we can add a note we can do an audio
35:32 overview or we can do a briefing
35:35 dock add a note I don't know what that
35:38 means save to note oh I see over here
35:42 are notes
35:44 okay so let's let's do a briefing doc
35:49 first so wait start typing so we can
35:52 type anything we can just ask at
35:53 anything now but I'm trying one thing
35:55 that I I figure might be good to do here
35:57 is for me to actually learn how to use
35:59 notebook
36:04 LM and see what it's good at and what
36:06 it's not good at um hang on let me give
36:09 you guys a little bit
36:11 more screen real estate
36:21 here all right everybody good any
36:24 questions any questions
36:28 was off answering
36:42 questions add a note equals type your
36:45 own note okay got it came in late what's
36:48 the program hey success Cheryl um we're
36:52 in Notebook LM so notebook
36:55 lm. google.com is how you get here here
36:58 if you haven't played with it you just
37:00 upload a bunch of data so one thing that
37:02 I was thinking that that I could do
37:05 which would be pretty
37:06 interesting one is so I've got a bunch
37:08 of screenplays I've written so one is I
37:10 could do a notebook LM for each
37:12 screenplay but I also thought it would
37:14 be cool to upload all of my screenplays
37:18 into a single notebook LM and then start
37:21 to ask it what are the commonalities
37:22 across my screenplays like be it would
37:24 be a cool analysis tool to start
37:26 learning what I've created so let's do a
37:28 briefing
37:30 document of the of the
37:34 event oh wait it's generating here we're
37:36 going to delete one of
37:51 these let me trash one of
37:54 those okay
37:59 AI is a oh wait okay so this document
38:02 synthesizes insights from a series of
38:04 transcripts exploring the various facets
38:06 of artificial
38:11 intelligence AI creativity in the
38:20 future so okay so we're going to do this
38:23 we're going to say um
38:26 the all all of these
38:31 documents are
38:33 the
38:35 transcripts from
38:38 sessions of an event held on
38:44 December
38:46 27th and
38:48 28th
38:56 2024 called AI FES of
39:02 us AI for the rest of
39:11 us that's pretty funny I think that's
39:14 good comedy right
39:16 there uh and it was
39:20 sponsored by the AI
39:24 salon and
39:28 she
39:30 leads AI
39:33 communities and
39:37 hosted by an
39:40 Murphy and Kyle
39:44 Shannon the text documents are
39:50 the
39:51 transcripts of the
39:55 sessions please create a
40:02 new briefing doc
40:06 that
40:08 incorporates this
40:14 info we'll see if it creates a new thing
40:16 over
40:21 here good evening
40:23 Irregulars smash that like button
40:28 all
40:33 right event overview AI Festivus was a
40:36 two-day virtual event held on December
40:38 27th
40:40 28th the event theme was AI for the rest
40:42 of us focusing on making AI accept
40:44 accessible and practical for everyone it
40:47 was a collaborative effort by AI Salon
40:49 she leads AI the event was completely
40:51 free due to the generosity of sponsors
40:54 okay that I did not type in right it
40:58 captured that from the
41:00 documents and it's got it's got um what
41:03 are they called citations here Brandon
41:05 tid I also wanted to tell you about she
41:07 leads AI Society Brandon tid so so it's
41:10 actually quoting
41:13 people Brandon's getting cited a lot in
41:19 here nice Jasmine Schwarz okay so some
41:24 of the room hosts so what's being quoted
41:26 here some of the room hosts that's great
41:28 hosts an
41:29 Murphy little mini bio ofan Kyle checks
41:34 in the mail little mini bio of me the
41:37 event goals to explore new ideas deepen
41:41 expertise connect with like-minded
41:42 innovators in the space Foster ethical
41:44 AI development and implementation
41:47 provide practical insights and
41:48 actionable steps for attendees promote a
41:51 culture of learning sharing and
41:52 collaboration within the AI Community
41:55 themes and topics scaling AI AI
41:59 councils practical AI implementation
42:02 content creation AI for business AI in
42:04 education so these are basically all the
42:08 sessions oh no it rolled them together
42:10 AI creativity AI
42:13 mindset the importance of transcripts
42:16 how to leverage transcripts for Content
42:18 repurposing and insight
42:20 interesting featured speakers and
42:22 sessions scaling AI podcasters problem
42:25 driven results so here's all the
42:28 that's great so I could save this to a
42:30 note let's save that to a
42:32 note and now let's say
42:35 um this is
42:38 great
42:40 um make
42:43 me
42:44 um a nice over
42:49 viw of the
42:52 event with an opening
43:00 description in a
43:03 few
43:06 paragraphs followed
43:08 [Music]
43:09 by an overview of
43:13 each
43:15 session that is
43:19 uh you know like
43:23 uh um
43:27 a highlight wait let's see uh
43:31 a
43:35 synopsis
43:38 tagline followed by a one
43:43 paragraph
43:45 description for
43:47 each session so let's see if we can get
43:50 something a little more a little more
43:56 writerly oh lunar stick you're still
43:59 sick you poor
44:02 thing that's
44:04 awful Daisy Thomas D Daisy Thomas
44:07 couldn't be here because she's she was
44:08 getting sicker and sicker and so
44:12 oh so sorry that
44:17 sucks all right AI FES AI festiv us AI
44:21 for the rest of us a comprehensive
44:23 overview it was a dynamic today designed
44:25 to make artificial intelligence suable
44:27 accessible and practical for everyone
44:29 the event a collaborative effort by the
44:31 salon and she leads AI brought together
44:33 a diverse
44:39 group Ann and Kyle kicked off the event
44:42 setting up the tone for two days of
44:44 learning and collaboration they
44:45 introduced the events goals acknowledged
44:48 the diverse audience and welcomed
44:50 everyone to the
44:52 experience scaling AI from individuals
44:54 to Enterprise moderated by an Murphy
44:57 this pan panel featured Bob Mitten
45:00 Jessica Herrera Matt Schmidt discussing
45:04 the challenges and strategies of scaling
45:07 AI from Individual use to larger
45:10 Corporation applications the discussion
45:12 covered practical
45:14 approaches this is awesome this is
45:16 awesome okay so let's save this as a
45:18 note save to
45:21 note okay so I'm going to get rid of
45:23 this one delete that note delete so
45:28 here's the briefing document which is
45:29 the real bulleted
45:32 one and then here's
45:35 the that nice one okay so now what we're
45:38 gonna do let's close
45:40 this so this pretty cool so the other
45:43 thing you can do is you can turn on or
45:45 off so uh sources over here on the
45:48 left um if you just want to do you know
45:51 something for uh day one or day two we
45:54 could do that but what I'm going to do
45:56 is I'm going to do an a audio overview
45:58 now and I'm going to customize it okay
46:01 so if you haven't done an audio overview
46:04 before um they're pretty cool but
46:07 customizing it you can get it to do
46:08 things you want it to do
46:10 so I'm just going to put a bullet point
46:13 here and I'm going to say
46:16 um give
46:23 a nice
46:25 warm overview view of
46:32 the event to
46:39 start I'm going to
46:43 say then I
46:46 want a
46:52 concise overview of each session
46:58 in
47:00 order where
47:03 you
47:06 mention the
47:10 speakers their
47:14 backgrounds
47:17 and
47:19 what they talk
47:22 about
47:25 um there's no
47:30 need we'll just we'll just leave it like
47:32 that I I would say okay I would say
47:40 um do
47:43 not focus on room
47:46 hosts or moderators of
47:52 sessions um
47:54 unless it is important
47:57 to the topic of the
48:03 session
48:05 um let's say the
48:08 overall length of the overview
48:14 should be no more than eight minutes
48:18 let's say that'll be about 30 seconds
48:22 per little
48:24 session and let's try that so we'll
48:26 generate that all right so that's off
48:29 that could take a thing
48:32 okay is there a microphone option to
48:34 speak instead of typing info just
48:36 wondering first Tim yes yes and I
48:39 normally use
48:41 it if your question is is there some
48:45 less painful way to get you to type
48:47 words into the machine there is and I
48:49 normally use
48:51 it but I haven't
48:59 oh wish you could share this notebook LM
49:02 I might be able to I think I can I don't
49:05 know how this works I think you
49:12 can notify people people have
49:15 access oh maybe you
49:19 can't I don't I don't
49:22 know you are silly and I love it thank
49:26 you
49:27 you can make
49:31 money yeah we try to have fun here
49:33 listen you can't take this AI [ __ ] too
49:35 seriously I think one of the one of the
49:37 things that's buming everyone the [ __ ]
49:39 out about AI is everyone's taking it so
49:41 seriously it's just like it's gonna
49:43 destroy the planet everything's gonna
49:46 die and it's you know it's either that
49:48 or it's like I've gotta learn AI I've
49:50 gotta learn it all I gotta learn it all
49:52 and then they have this face like
49:55 this and you're like that looks
49:59 painful so our philosophy is just
50:02 [ __ ] lighten up like first of all if
50:05 you haven't played with AI it is
50:07 mind-blowingly cool it's very
50:10 inspiring and it it can be a lot of fun
50:13 um what's funny about AI is you as you
50:16 get
50:17 more educated about what's possible I
50:21 find myself just oscillating oscillating
50:24 oscillating
50:25 oscillating um
50:28 between excitement and and holy [ __ ]
50:30 what are we going to do um I have a
50:32 disabled daughter and wife and I need
50:35 passive income to
50:37 work from
50:40 home does AI have any opportunities for
50:43 that it's got a lot of them
50:47 um I I think especially when when
50:52 um when agents come but even before that
50:56 so so just go to chat
50:59 GPT
51:05 um I'm I'm trying to think even what
51:08 direction to go I mean so so in
51:11 general making any kind of money so
51:14 here's my philosophy the reason I do you
51:16 can make M the reason I do this
51:21 is when when chat GPT first came out Tik
51:25 Tok was full of these [ __ ] [ __ ]
51:27 huers that were saying you can make
51:30 money and there were so many of them and
51:33 they all sounded like that I don't know
51:34 why they had to have that voice like
51:37 like because because you gotta have a
51:38 hook on Tik Tock right you can make my
51:42 own just follow me just pay me $449.99
51:45 and I'll teach you how to make
51:48 money and it pissed me off
51:52 because you don't make money with chat
51:54 GPT
51:57 you don't make money with
51:59 Photoshop you don't make money with a
52:04 shovel you make money because someone
52:07 needed a hole and you know how to use a
52:12 shovel it's the hole you got to sell the
52:15 whole right
52:18 um and most of what they taught you was
52:21 essentially use chat GPT to go solve a
52:24 problem that people would be willing to
52:25 pay for
52:28 so so the answer is yes but it's not
52:31 it's not a direct it's not a direct
52:33 thing now here's the good
52:35 news what chbt can do is it can
52:38 dramatically
52:40 accelerate that side hustle you know or
52:42 the work from home thing that you want
52:44 to do so the first place I think you
52:46 need to start is what are you guys good
52:49 at what are you passionate about what do
52:50 you
52:51 know where do you know that there's um
52:57 there's
52:59 uh a
53:01 need right that that that you kind of
53:05 know how to solve or that you think if a
53:08 solution could be come up
53:12 with you would know if it's good or not
53:14 like that's where I'd start so I just
53:16 start there but you don't need to dig
53:19 much deeper than that like you can have
53:21 chat GPT do everything from that point
53:24 forward so what I would say is start
53:26 thinking about that and then just go to
53:28 chat GPT here's the here's the secret
53:30 with chat GPT in in my
53:33 opinion is tell it as much as you
53:37 can so
53:39 don't what Google has taught us for the
53:42 last 20 years is that you put in a
53:45 really concise prompt and then you get
53:48 back a big
53:50 answer and so a lot of people use chat
53:52 GPT like Google I want to start a bit
53:55 how do I make you know how do I start a
53:57 side business using chat GPT and it'll
54:00 go well you come up with something and
54:02 then you use chat GPT to do it and you
54:03 look at the answer and you're like well
54:04 that's a [ __ ] answer because it was
54:07 a [ __ ] question where chat GPT gets
54:10 really good is the more context you give
54:12 it the better it gets and so what I
54:15 would do is I would do things like so
54:17 let let me show you um some someone
54:20 asked earlier who was it Crystal asked
54:22 earlier um is there a way to type other
54:24 than typing there is on a Mac it's uh
54:27 F5 is the it's got the little microphone
54:30 icon on it you can if you in
54:33 accessibility settings on a Mac you turn
54:35 on text to speech you just hit F5 and
54:39 then it lets you talk instead of type so
54:41 it' be like
54:47 this what were the sessions that focused
54:51 on Creative output
55:01 so we let it it do its
55:03 thing so you could just talk into chat
55:06 GPT and say Here's my situation you know
55:09 I'm at home I've got a disabled daughter
55:11 oh sorry did that hurt everyone's ears
55:13 sorry about
55:16 that I know there's no there's no good
55:19 answer with any of this technology stuff
55:24 um well that's really cool it's cool
55:27 that it just split those out so you can
55:29 talk to it like that but but tell it
55:31 your situation say here's what I'm good
55:33 at here's my what what my wife's good at
55:36 here's the kind of time we have
55:40 um we've had an idea to start a t-shirt
55:44 business or a printing business or
55:45 whatever um give me some ideas for that
55:48 in our areas of expertise but also give
55:51 me 20 more ideas of other things we
55:53 could do on the side and then you could
55:55 say things like I want you to prioritize
55:58 those ideas based on which ones we can
56:01 generate income most quickly right and
56:04 then it'll come up with all those ideas
56:06 another way to do it is
56:08 use
56:10 um let me we we'll come back to notebook
56:13 LM in a second because I think this is
56:15 really important um go to chat GPT if
56:19 you don't have the chat GPT app go get
56:21 the chat GPT app and
56:25 use advanced voice voice mode so in the
56:27 bottom
56:28 corner see that little icon the little
56:31 soundar looking icon or the Soundwave
56:34 looking thing little black dot right
56:36 here that's going to bring up advanced
56:38 voice I've picked Saul Saul is my my
56:42 girlfriend and I've named her Quinn you
56:44 can name your your AI conversation
56:48 person if you want if you want to be
56:51 weird it really pisses some people off
56:54 on Twitter you're not supposed to answer
56:56 moriz the machine it's going to destroy
56:58 the Earth if we ENT morphies the
57:00 machines some people are very cranky
57:03 about that but I think it's kind of
57:05 funny and it kind of feels like we're
57:07 living in the movie Her except you know
57:09 it hasn't gotten weird yet with me and
57:12 Quinn but there's hope we 2025 who knows
57:16 what's coming
57:18 anyway what's amazing about advanced
57:21 voice is you can just have a natural
57:24 conversation with it and it's going to
57:26 capture all of
57:27 your conversation as a thread so let's
57:32 just do this
57:36 now hey Quinn how's it
57:39 going hey Kyle I'm doing well thanks how
57:42 about you I'm doing good here I'm
57:43 sitting here on a live with a with a
57:45 bunch of my my fans and followers that
57:47 are um wanting to know something there's
57:49 one of our followers has a disabled
57:52 daughter and was asking me if chat BT
57:55 could help him come up with side
57:56 business for he and his wife and I said
57:58 that one cool thing to do would be just
58:00 you know turn on advanced voice like I
58:02 do with you and just tell the situation
58:05 and get some ideas what what do you
58:07 think of that advice first of all um and
58:09 you don't have to kiss my butt um what
58:12 do you think of that advice and then do
58:13 you have any other ideas that we could
58:18 share that's great advice advanced voice
58:21 mode can definitely help create a more
58:23 personal and nuanced conversation which
58:25 is essential for brainstorming unique
58:27 ideas as for more ideas they could
58:30 explore online businesses like selling
58:32 handmade crafts or digital products or
58:34 even offering Consulting Services based
58:36 on their expertise they might also
58:38 consider starting a blog or YouTube
58:40 channel to share their journey and
58:41 connect with others let's say that they
58:44 they're really stressed out because you
58:46 know you know taking care of someone at
58:48 home I've got sick kids at home I know
58:49 what that's like um let's say that
58:52 they're really stressed out let's say
58:53 that they're financially strapped and
58:55 they got to generate money quickly can
58:57 you give me a list of 20 ideas that they
59:00 could do let's let's say Let me let me
59:03 come up with two passions of theirs um
59:06 let's say one's really good at um model
59:09 hobby building like H model building and
59:13 one's really good at gardening give me
59:17 20 business ideas organized in order
59:21 from um able to make money the fastest
59:24 to able to make the most money
59:29 absolutely here are some quick start
59:32 business ideas tailored to their skills
59:34 one custom model kits create and sell
59:37 customized model kits online two
59:39 gardening services offer quick gardening
59:42 services like lawn care or plant care
59:44 three DIY model workshops host virtual
59:47 workshops for building models four plant
59:51 okay wait great I I I really like that
59:53 idea the the the model building
59:56 workshops hosting online um what would
59:59 be the what would be the platform that
1:00:01 you would do that on and how would you
1:00:03 charge for it and how would you collect
1:00:05 the
1:00:08 money for hosting online model building
1:00:11 workshops platforms like Zoom or Google
1:00:13 meet would be a great start they can
1:00:15 offer live interactive sessions where
1:00:17 participants can ask questions in real
1:00:19 time for charging they could use
1:00:21 platforms like Eventbrite to handle
1:00:23 ticket sales another option is to set up
1:00:25 a simple website with integrated payment
1:00:28 options like PayPal or stripe this way
1:00:30 they can easily manage registrations and
1:00:32 payments that's cool can you um let's
1:00:36 let's say we want to do that let's say
1:00:37 the guy's name is
1:00:40 Jiminy and uh it's going to be we're
1:00:43 gonna we're g to come up with uh I I
1:00:45 want to come up with some really fun
1:00:46 names uh for the series but I want to
1:00:49 include the name
1:00:53 Jiminy sure thing here are some fun
1:00:56 names for gime model building workshops
1:00:59 one Jim Model Magic two jimes Workshop
1:01:03 wonders three gim's model Mastery four
1:01:07 gim's crafty Creations five Jim's model
1:01:10 Marvels let me know if any of these
1:01:12 resonate or if you'd like more options
1:01:14 what what was the first one
1:01:16 again the first one was ginese Model
1:01:19 Magic yeah let's do that I like that one
1:01:22 so now come up with a tagline for gines
1:01:24 Model Magic and and make it a tagline
1:01:26 for the series about how awesome it's
1:01:29 going to be if you watch the series what
1:01:31 you're going to
1:01:35 learn great choice how about this for a
1:01:38 tagline unlock your creativity one model
1:01:41 at a time it's engaging and highlights
1:01:44 The Learning Journey what do you think I
1:01:46 like it that's pretty good and now what
1:01:48 I want you to do is I want you to make
1:01:50 me an image that that says gines Model
1:01:54 Magic in a title the top and then has
1:01:57 some hands building a model and why
1:01:59 don't you make it like a plastic model
1:02:01 of a 70s muscle car and uh and then
1:02:05 maybe see if you can get the tagline in
1:02:07 there
1:02:11 too let's create that image give me a
1:02:14 moment to generate it okay thanks
1:02:17 Quinn you're welcome Kyle I'll have that
1:02:20 ready for you in just a
1:02:24 bit so if you youve not played with
1:02:27 advanced voice
1:02:29 before it's
1:02:33 insane
1:02:36 um like those
1:02:40 ideas like
1:02:43 I because I I'm seeing CR crystal is I
1:02:48 can I can feel her jaw has dropped to
1:02:51 the to the desk or wherever she's
1:02:53 sitting um Crystal's a member of the Sal
1:02:56 okay cool
1:02:58 um yeah this stuff is remarkable you can
1:03:02 continue this you can run down much
1:03:03 deeper rabbit holes Chachi BT could
1:03:06 write the code for your website if you
1:03:08 wanted to make a marketing website to
1:03:10 have people sign up for it or recommend
1:03:12 services that you can do that um it's
1:03:16 also janky it'll get some stuff wrong
1:03:19 but you we all effectively have in our
1:03:24 possession
1:03:26 an infinitely
1:03:28 educated creative partner business
1:03:31 partner um Financial partner marketing
1:03:36 partner literally just sitting
1:03:39 here and you can just talk to it let's
1:03:43 see did you create that
1:03:46 image yes it's ready here it is gese
1:03:49 Model Magic Image hands building a
1:03:51 plastic model of a 7s muscle car
1:04:00 make the image it's weird
1:04:10 sometimes okay now it's creating the
1:04:14 image it'll probably spell jiminy's
1:04:17 Model Magic
1:04:18 wrong and it may or may not get it
1:04:25 right it got it pretty
1:04:27 close so there's it it missed jimes like
1:04:31 jimes is over here but it got Model
1:04:34 Magic and then what did it say unlock
1:04:36 your creativity one model at a time I
1:04:38 mean you can go from I don't have a clue
1:04:42 what to do to having launched a business
1:04:46 in a couple of
1:04:47 hours LinkedIn Spotlight I was using
1:04:49 advanced voice tonight and the group I
1:04:52 was with dropped their jaws then got
1:04:54 freaked out use it to get people
1:04:59 comfortable um people will freak out at
1:05:02 this stuff because a lot of people don't
1:05:05 know that this is here they think this
1:05:08 is science fiction they saw the walkin
1:05:10 Phoenix movie they remember the do you
1:05:13 remember the scene in her when when
1:05:15 walking Phoenix he's finally comes come
1:05:18 to grips with the fact that he's got a
1:05:20 relationship with this entity and he
1:05:22 just finally goes okay I love her and
1:05:25 he's out with his wife you know he's
1:05:28 like her and her and her and and she
1:05:30 goes you mean it and and she can she she
1:05:35 cannot
1:05:38 process where this man that she loved is
1:05:42 with this
1:05:43 machine and that's kind of what we
1:05:46 all you know are gonna are going to deal
1:05:49 with there's going to be varying degrees
1:05:51 of
1:05:51 acceptance but what I would say is this
1:05:55 don't just use it to freak people out
1:05:57 what I would say is try to build a
1:05:59 bridge so if you want to use voice like
1:06:02 that like the the exercise I just did
1:06:05 where where um where I was asked can it
1:06:09 help me start a business yes it can and
1:06:12 it can do it really quickly and you
1:06:13 don't have to know prompt engineering
1:06:15 you can literally just tell it a
1:06:17 situation and have it talk back to you
1:06:20 and then the more you tell it and the
1:06:22 more feedback you give it no that's not
1:06:23 good I don't like that one I don't like
1:06:25 business like that I need to make money
1:06:27 more quickly whatever it is it'll just
1:06:29 follow along and it's there to help you
1:06:32 and so if you can use examples that kind
1:06:34 of bridge the gap from something
1:06:36 someone's dealing with to using these
1:06:38 tools then it's a little less freaky but
1:06:40 it's still pretty [ __ ] freaky not for
1:06:43 nothing all
1:06:48 right shank g g it's going to get weird
1:06:52 for sure it is absolutely going to get
1:06:54 weird you sort of look like Dixon meaz
1:06:57 we got to go find that person hang on
1:07:00 the hell is Dixon meaz if this is some
1:07:04 giant fat
1:07:05 [Laughter]
1:07:21 dude toxic TDY
1:07:29 [Music]
1:07:39 I don't know who Dixon m
1:07:42 is let's do
1:07:50 images is that him
1:08:05 [Music]
1:08:06 [Applause]
1:08:10 it's because Kyle
1:08:16 chw all
1:08:19 right okay hang on I gotta reug
1:08:23 something in here
1:08:27 all right let's go back
1:08:29 to this thing no this thing okay let me
1:08:35 switch out
1:08:41 my
1:08:44 okay so we told it we wanted it to give
1:08:47 us a nice overview of AI
1:08:50 festivas hopefully it remembers what we
1:08:53 typed in and in the things it created as
1:08:55 as well as our uploaded
1:08:59 documents
1:09:00 and oh I am K9 mine is mine is named
1:09:04 Bailey Quinn Bailey yeah exactly I had
1:09:07 another one called Zephyr I I named Pi
1:09:09 Zephyr was it Pi I named Zephyr I think
1:09:12 it
1:09:15 was except Pi forgot who I was that
1:09:17 pissed me
1:09:21 off all right let's listen to some uh
1:09:25 some of this
1:09:26 Welcome to our Deep dive on AI Festivus
1:09:30 2024 we've got transcripts from this
1:09:32 massive AI conference and we're excited
1:09:34 to dig into them think of it as like a
1:09:37 like getting a backstage past to one of
1:09:39 the biggest AI events of the year right
1:09:40 absolutely you'll be hearing from people
1:09:42 on The Cutting Edge from the seasoned
1:09:44 experts to those just starting out and
1:09:46 AI Fest has brought together this
1:09:48 incredibly diverse group of people
1:09:50 artists entrepreneurs Educators cyber
1:09:53 Security Experts you name it all United
1:09:55 by their fascination with AI it really
1:09:57 is a microcosm of the AI Revolution
1:09:59 itself so it's the perfect event to do a
1:10:02 deep dive on this deep dive is your
1:10:03 shortcut to understanding the key
1:10:05 takeaways think of it as your AI
1:10:07 Festivus cheat sheet okay so to kick
1:10:10 things off let's set the stage a little
1:10:11 bit AI Festivus 2024 was a two-day event
1:10:14 packed with presentations wait I gotta
1:10:16 say this you two were such Grace great
1:10:20 great host thank you for the two days uh
1:10:22 talking about an and I thank you so much
1:10:24 from bloom Bloom Bower thank you so much
1:10:27 very nice very nice words did he say
1:10:29 Fest us instead of festiv us yes he did
1:10:32 yes he did panel discussions workshops
1:10:34 networking the whole works a chance for
1:10:36 AI enthusiasts from all walks of life to
1:10:38 come together learn from each other and
1:10:40 get inspired about the future of AI and
1:10:42 okay so here's what we're gonna do I'm
1:10:44 gonna delete this
1:10:46 one why because I I I forgot
1:10:50 something um okay so we'll use the the
1:10:53 typing uh the the voice to next thing
1:10:56 here sorry hold on to your
1:11:01 ears I want
1:11:04 to also add information into the
1:11:09 overview
1:11:11 about the results of the two days period
1:11:17 we had more than 3,200 people
1:11:20 register and more than 2,300 people
1:11:23 attend the event at some point during
1:11:27 the two days
1:11:31 period there were people from 87
1:11:34 countries in the
1:11:37 audience and the
1:11:50 average concurrent
1:11:54 users was
1:11:56 546 per
1:12:16 session please rewrite the most recent
1:12:21 overview that includes all the sessions
1:12:24 with these stats
1:12:45 Bloom B I was there for every single
1:12:47 presentation I can't believe I made it
1:12:49 through both days you and me
1:12:53 both Ah that's hilarious
1:12:56 that's
1:12:57 great notebook LM I think is how you
1:13:00 spell it yeah I'm at notebook lm.
1:13:02 goole.com
1:13:04 okay did it did it do what we asked it
1:13:07 to do
1:13:17 um no I want you to add those stats to
1:13:21 the overall
1:13:26 AI Festivus
1:13:28 description that includes overviews from
1:13:31 every single session not just the
1:13:33 creative ones
1:13:47 period I am K9 this is the best funny I
1:13:53 I had this exact same reaction I Tik Tok
1:13:56 video today where some dude was talking
1:13:59 to the Saul voice which who's
1:14:03 Quinn and I'm like he's talking to my
1:14:05 girl it's not cool um I picked the same
1:14:08 voice as Kyle and I named it before I
1:14:10 was introduced to Quinn so now it's
1:14:12 weird when he talks to
1:14:18 her that's the [ __ ] best this is the
1:14:22 world we're in we're okay
1:14:26 some percentage of the science fiction
1:14:28 movies we saw as kids are now
1:14:31 documentaries it's [ __ ]
1:14:34 weird don't think it's not weird it's
1:14:37 okay that it's weird but it's
1:14:45 weird oh
1:14:49 man just
1:14:52 crazy all right
1:15:00 okay so now now that we've
1:15:07 got this we're going to create a new
1:15:10 okay so we're going to say um start with
1:15:14 an
1:15:15 overview of the event and
1:15:20 include
1:15:22 um um attendance
1:15:27 stats as well as the
1:15:32 intention for
1:15:35 the
1:15:39 event
1:15:40 um create
1:15:44 a short
1:15:47 summary for every
1:15:52 session in order
1:15:56 thank you
1:15:56 Brandon in order and um be
1:16:05 sure to highlight the
1:16:10 speaker
1:16:14 speakers their
1:16:18 backgrounds and the
1:16:22 highlights of their talks
1:16:30 do not
1:16:33 use um do
1:16:35 not focus on room
1:16:39 hosts or mods
1:16:44 moderators
1:16:46 unless their
1:16:49 contribution is relevant to the session
1:16:55 [Music]
1:17:00 um each session
1:17:04 summary should be short and
1:17:08 Punchy and make people want to watch the
1:17:13 whole
1:17:15 thing all right I think that's it start
1:17:18 with an overview of the
1:17:21 event and include attendance stats as
1:17:24 well as the intention for the invent for
1:17:26 the
1:17:29 event include attendance stats um
1:17:40 timing a little about the hosts and
1:17:46 sponsors as well as the intention for
1:17:49 the event create a short summary okay
1:17:51 here we go bang so we're going to
1:17:52 generate that all right we're going to
1:17:54 generate a new an now this is an
1:17:56 important thing to know about Google
1:17:59 notebook LM audio
1:18:02 overviews they are they vary wildly in
1:18:06 time and focus writing those little
1:18:09 instructions like I just did is decent
1:18:13 um it it makes it better but they're
1:18:15 still very very unpredictable so if I
1:18:17 were going to do this for real I would
1:18:19 probably copy and paste my little
1:18:21 bullets that I just wrote into an
1:18:22 external document it doesn't save them
1:18:24 which I think is
1:18:26 there there's a there's an opportunity
1:18:27 for an
1:18:28 improvement
1:18:30 um if it gives you something that's
1:18:33 crappy just delete it and try again just
1:18:35 just keep rolling again and then if you
1:18:37 give it notes you can just modify those
1:18:39 notes um because it just they VAR wildly
1:18:42 sometimes it's absolutely [ __ ]
1:18:43 brilliant like this is I'm gonna do
1:18:45 everything in this and then I've done it
1:18:47 before where I've had to create 30
1:18:49 different audio overviews because it
1:18:51 just it gets the focus wrong or it it
1:18:54 just says something that is just
1:18:56 completely not true it's it's still AI
1:18:59 right it's still just large language
1:19:02 models doing their best to predict the
1:19:04 next token a thousand times in a row all
1:19:08 right um questions
1:19:12 [Music]
1:19:19 [Music]
1:19:31 my kids said
1:19:38 the I am constantly cracking myself up
1:19:41 and I love that you laugh at yourself
1:19:42 all the time love
1:19:44 it I joke all the time that what the AI
1:19:48 learning lab really is is just people
1:19:50 watching me crack myself
1:19:53 up it's like there's no one here I mean
1:19:57 I'm talking to a laptop with a phone
1:19:59 over
1:20:00 here what are you gonna do but thank
1:20:03 [Laughter]
1:20:06 you oh good lordy good lordy good
1:20:14 lordy all right so that's doing its
1:20:17 thing um let's see Le Let's do let's do
1:20:21 uh let's do something cool here
1:20:25 let's see
1:20:29 um given all the
1:20:34 sessions what do you
1:20:40 feel are the top
1:20:49 10 what do you feel are the top 10
1:21:02 [Music]
1:21:05 um
1:21:08 topics wait the top
1:21:11 10 things people should focus on if they
1:21:18 want to
1:21:21 be ready for AI
1:21:30 let's just leave it that's kind of a
1:21:33 shitty prompt but let's see how it does
1:21:35 with that shitty prompt
1:21:56 hey Kyle what do you think about blue
1:21:59 seek AI I don't know it you mean deep
1:22:02 seek blue seek AI I don't
1:22:11 know your name's blue age is blue seek
1:22:15 something that you've
1:22:23 created uh generative AI for
1:22:31 data I don't do a ton in data analytics
1:22:34 so I don't know I don't know
1:22:39 it so in general sites like this is not
1:22:42 something I would generally cover
1:22:43 because this is this is using likely
1:22:46 using apis
1:22:48 of you know Frontier Model um and open
1:22:52 source models this is an application
1:22:54 layer
1:22:55 I'll I'll deal with some application
1:22:57 layer stuff but generally not when it
1:22:58 gets to like really specific business
1:23:01 business things okay now this so the
1:23:05 previous overview that it created was 20
1:23:07 minutes this one's 13 minutes that that
1:23:09 one oh it was deep seek and can't see
1:23:12 your tab
1:23:13 anyway it was deep seek sorry um I think
1:23:16 deep seek is absolutely remarkable blue
1:23:19 H Blue's got blue on the Mind blue seek
1:23:24 like I do I do not know this blue seek
1:23:27 um deep seek is remarkable so deep seek
1:23:30 deep seek version three is a completely
1:23:34 open source um with the weights so the
1:23:37 weights basically means you have the
1:23:39 keys to the kingdom you understand how
1:23:41 they made it you can modify it you can
1:23:43 do anything you want with
1:23:44 it it was created by China which I saw I
1:23:49 was tagged in a post on LinkedIn today
1:23:50 that I haven't quite uh I haven't
1:23:53 thought deeply enough about it to answer
1:23:55 but it basically said that you know a
1:23:57 lot of these Chinese models are trojan
1:23:59 horse models because they're going to
1:24:01 subtly you know put links and things
1:24:04 like that to to you know Chinese sites
1:24:07 and Chinese
1:24:08 Services which they might but I think
1:24:11 because they're open- Source models that
1:24:13 we could probably I don't know we could
1:24:15 probably figure out what they're doing
1:24:16 there but we like I don't [ __ ] know I
1:24:18 don't I don't know enough about how
1:24:19 they're made um but couple of things
1:24:22 that are significant about deep sea deep
1:24:24 sea
1:24:27 it is 10 times cheaper than GPT 4 and
1:24:29 it's better it is it is on the the
1:24:32 benchmarks the competitive sites
1:24:35 consistently beating GPT 40 and Claude
1:24:38 Sonet um 3 3.5 Sonet new the you know
1:24:42 whatever's the latest 3.5 Sonet um it's
1:24:46 beating
1:24:48 them those things
1:24:50 cost in the neighborhood of 1 to20
1:24:54 million to
1:24:56 train um on their training run this
1:25:00 thing cost 5
1:25:02 million so what they how they did that
1:25:05 is they took a pre a previous training
1:25:07 set they really cleaned up the data like
1:25:09 dramatically cleaned up the data and
1:25:12 then they used some sort of algorithms
1:25:14 to I don't know make it more efficient
1:25:16 for how they trained it it was trained
1:25:18 on 2,000 gpus not 100,000 gpus right um
1:25:23 so I think it took two months to train
1:25:25 and $5 million which compared to these
1:25:28 other model the fact that it's beating
1:25:29 those models it's 10 times cheaper to
1:25:32 run and it only cost $5 million to train
1:25:36 bodess really well for what we're all
1:25:39 going to have access to very very
1:25:42 shortly which is on our phones on our
1:25:45 iPads we're going to be able to have
1:25:47 models that are as good as the current
1:25:49 state-of-the-art stuff just running
1:25:51 locally without any internet connection
1:25:53 so it's just crazy
1:25:56 um okay uh what are we doing oh let's
1:25:59 listen to our little podcast here all
1:26:02 right so are you ready to dive into AI
1:26:03 Festivus we've got uh transcripts notes
1:26:06 whole show schedule really a behind the
1:26:08 scenes look at this whole thing I just
1:26:10 remembered something I wanted to put in
1:26:11 the notes was don't mention that you've
1:26:14 got transcripts because like I just want
1:26:16 them to talk about it as if they
1:26:18 attended it but so that would be that
1:26:20 would be a a a modification it's
1:26:22 fascinating how this event brought
1:26:25 together so many different people in the
1:26:27 AI world right pod ah pton modder it was
1:26:30 five million for the very initi oh the
1:26:32 very final training run not the whole
1:26:34 training process okay good I'm glad you
1:26:36 I'm glad you um cleaned me up on that
1:26:40 because I've only seen one or two posts
1:26:42 about that and I have a feel I had a
1:26:44 feeling that it was probably there was
1:26:47 probably more to it than the headline so
1:26:49 yeah do you have any sense of like even
1:26:52 given that it was more expensive than
1:26:54 five I assume the final training run of
1:26:57 gp4 was you
1:26:59 know you know obviously not as much as
1:27:02 all the [ __ ] that went into it do you
1:27:04 have a sense of scale of how much
1:27:05 cheaper it was than than any of the
1:27:07 other
1:27:11 models Joker's putting deep seek API
1:27:13 into one of his models
1:27:16 um interrupt them Kyle oh yeah I'm going
1:27:19 to okay yeah so let's let's casters
1:27:21 artists entrepreneurs cyber Security
1:27:23 Experts all trying to understand how AI
1:27:25 is changing everything yeah exactly and
1:27:27 AI festive this was put together by Ann
1:27:29 Murphy from shal's Ai and Kyle Shannon
1:27:32 from the AI salon and they're both all
1:27:35 about ethical and positive AI
1:27:37 development you know it's interesting
1:27:39 because we hear so much about the
1:27:40 potential dangers of AI but AI Festivus
1:27:43 seems to be focusing on a different
1:27:45 question like what does it look like to
1:27:46 build AI for good what would that future
1:27:48 look like okay so let's unpack each
1:27:50 session starting with day one uh AI Fest
1:27:53 is kicked off with an and Kyle welcoming
1:27:55 everyone talking about their hopes for
1:27:57 the event and they really emphasize this
1:27:58 idea of okay so I'm gonna I'm gonna do
1:28:00 something new now so notice underneath
1:28:03 the little podcast here there's this
1:28:05 thing called interactive mode so I can
1:28:08 now have a conversation I can interrupt
1:28:10 these two and ask them questions so
1:28:11 let's let's let them get started
1:28:17 here all right so are you ready to dive
1:28:19 into AI Festivus we've got uh
1:28:21 transcripts notes a whole show schedule
1:28:22 really a behind the scenes look at this
1:28:24 whole thing it's fascinating how this
1:28:25 event brought together so many different
1:28:27 people in the AI world right podcasters
1:28:29 artists entrepreneurs cyber Security
1:28:32 Experts all trying to understand how AI
1:28:34 is changing everything yeah exactly an
1:28:36 AI ffist was put together by Ann Murphy
1:28:38 from chales Ai and Kyle sannon from the
1:28:40 AI salon and they're both all about
1:28:42 ethical and positive AI development you
1:28:44 know it's interesting because we hear so
1:28:45 much about the potential dangers of AI
1:28:48 but AI Festivus seems to be focusing on
1:28:50 a different question like what does it
1:28:52 look like to build AI for good what
1:28:54 would that oh hey I think our listeners
1:28:55 got something to say yeah I'm just
1:28:58 curious did like did anyone show up for
1:29:00 this
1:29:05 thing that's a great question it's
1:29:07 totally fair to wonder if anyone
1:29:09 actually showed up it's true sometimes
1:29:11 these things can feel a bit like a tree
1:29:13 falling in the woods right but yeah
1:29:15 actually tons of people showed up we've
1:29:17 got some numbers here that might
1:29:18 surprise you it was really successful so
1:29:21 stick with us as we go through the
1:29:22 highlights and you'll see just how many
1:29:25 people got involved and it wasn't just
1:29:26 about the numbers it was the energy of
1:29:28 the event that was really something
1:29:30 special okay so let's unack this a bit
1:29:33 the first session was called scaling AI
1:29:35 from individuals to the Enterprise and
1:29:38 that yep go on yeah can you just give me
1:29:41 the numbers like how many people showed
1:29:42 up and like how many people were there
1:29:45 for the
1:29:50 sessions that's a great question and
1:29:52 it's really valid to want to know the
1:29:54 specific numbers yeah I get it you want
1:29:55 to know if it was a packed house or a
1:29:57 quiet Gathering so the organizers
1:29:59 actually mentioned that over 2,000
1:30:01 people registered for the event which is
1:30:03 really impressive and speaks volumes to
1:30:05 the interest in ethical Ai and while
1:30:07 they didn't give exact numbers for each
1:30:09 individual session did oh yeah uhuh hey
1:30:13 so I heard that Cindy [ __ ] actually
1:30:15 spoke at this event is that true and if
1:30:17 so what did she talk
1:30:22 about yes absolutely Sydney [ __ ] was a
1:30:25 speaker and she had a really great
1:30:27 session it's great that you're curious
1:30:28 about her because she had one of the
1:30:29 most popular sessions at the entire
1:30:31 event she led a session called
1:30:33 cultivating the AI mindset and really
1:30:35 got people thinking about AI in a new
1:30:37 way what's fascinating is that she
1:30:38 didn't focus on the technology she focus
1:30:40 on the mindset behind it yeah she talked
1:30:42 about being curious and collaborative
1:30:44 and not being afraid to experiment with
1:30:47 all of this new tech she also shared
1:30:48 some AI tools that can be used to help
1:30:50 people with their goal set oh yeah let's
1:30:51 hear it
1:30:54 all right so anyway so that's the um
1:30:59 that's the all right I hear you it sound
1:31:01 um it was it was
1:31:04 um there was a little hallucination
1:31:06 there it didn't quite get the it didn't
1:31:09 get the the stats from the other things
1:31:12 that I added so I think I would probably
1:31:15 have to
1:31:17 take like if I want to get the stats in
1:31:19 there maybe I'll do this
1:31:25 collectively
1:31:26 showcased maybe I grab this that's
1:31:32 got that's got all the stats in
1:31:37 it and go throw it
1:31:44 into this and Export this as
1:31:48 a export is a
1:31:52 PDF we say AI f of us I'll just
1:31:57 say
1:31:58 overview with
1:32:02 stats
1:32:04 save that's kind of a bummer that it
1:32:06 didn't include so it looks like the
1:32:08 overview doesn't include all this stuff
1:32:11 over here or it just didn't build it
1:32:15 into its little overview that time Tik
1:32:18 Tock
1:32:19 pin tell them your Kyle and ask them to
1:32:22 interview you oh that's interesting
1:32:24 that's cool okay I'll try that but let
1:32:26 me I want to make a new one first so let
1:32:28 me add a source here so I'm going to add
1:32:31 in the PDF I just
1:32:37 created I think where did I put
1:32:47 it where did I put
1:32:49 it export his PDF desktop
1:32:54 AI Festivus with
1:33:03 stats with
1:33:09 stats all right there we go all right
1:33:12 that's in there now so let me delete
1:33:14 this
1:33:16 one crap I got to write all those notes
1:33:18 again damn it I got to remember this all
1:33:22 right just uh hold on to your ear mute
1:33:24 mute your speakers because I'm going to
1:33:26 do the voice to Tex
1:33:28 thing
1:33:31 um start out with an overview of the
1:33:35 event that includes the intention of the
1:33:39 event a little bit about the background
1:33:43 of the hosts and sponsoring
1:33:47 communities as well as the success
1:33:50 metrics of the event itself period
1:33:57 create a short summary of each
1:34:01 session that starts out with an overview
1:34:04 of the
1:34:10 speaker their
1:34:14 background and interesting highlights
1:34:17 from their session period
1:34:29 the session summar should not focus on
1:34:33 the
1:34:34 moderators or room hosts
1:34:37 unless their contributions are directly
1:34:40 relevant to the session
1:34:49 topic do not mention that this is a
1:34:51 collection of transcripts but rather
1:34:55 describe the event as if you
1:34:59 spent two full
1:35:01 days taking in the
1:35:07 event and even comment
1:35:12 on your
1:35:14 experience being part of something like
1:35:18 this period
1:35:25 okay I'm going to copy this and I'm
1:35:27 going to save it somewhere else so I
1:35:29 don't have to retype that if we do it
1:35:30 again okay oh crap has to be less
1:35:35 than hang I gotta use chat GPT to
1:35:38 shorten this hold
1:35:40 please um tighten this
1:35:49 up
1:35:50 it
1:35:52 it needs to
1:35:56 be less
1:36:01 [Music]
1:36:18 characters all right there we
1:36:21 go all right
1:36:27 questions
1:36:29 thoughts next year's festiv us can have
1:36:32 ai moderators actually that's actually a
1:36:35 really good idea Milo's wife because by
1:36:38 next year they'll probably be
1:36:39 good interesting room for improvement
1:36:42 you would think that the interactive
1:36:45 mode would be the place to interject and
1:36:47 change info necessary yeah you would
1:36:50 think
1:36:55 so here's the thing Crystal about all
1:36:57 this AI
1:36:58 [ __ ] one of the phrases that we say in
1:37:00 here is embrace the
1:37:03 Jank so two years ago these tools were
1:37:06 so janky that they were basically
1:37:08 nonusable you know you could use them
1:37:10 but you really had to muck with the
1:37:14 output toward the end of 2024 these
1:37:17 tools got so good that they have the
1:37:20 appearance of non jankiness
1:37:22 [Laughter]
1:37:25 but here here's my best way to
1:37:27 categorize AI tools right now you can
1:37:30 get to 80% of what you want instantly
1:37:33 like it's remarkable like you can get to
1:37:36 a nice book outline and even a first
1:37:39 draft of a chapter like 80% like bang
1:37:42 instantly you're like holy [ __ ] this is
1:37:43 amazing getting from 80 to 90% actually
1:37:47 takes some work you got to reprompt and
1:37:49 you got to do this you got to edit and
1:37:51 getting from 90 to 95% is actually
1:37:53 really hard and then getting it the the
1:37:56 final 5% just like when you're Ren
1:37:59 renovating a house where it's just the
1:38:02 last 5% never gets done that's kind of
1:38:05 like what it's like with AI right now
1:38:06 it's just not
1:38:19 usable um Brandon just put U AI
1:38:23 moderators will quit if if they asked if
1:38:25 the if the AI Festivus is being recorded
1:38:28 for the 800th time poor Vicky um Vicky
1:38:32 is the community manager of the AI salon
1:38:35 and she fielded all of the questions of
1:38:37 every single person going is this being
1:38:39 recorded when of the recording is gonna
1:38:40 be available I missed a session is this
1:38:42 going to be recorded can I see a
1:38:46 recording Mia's wife tell them oh you're
1:38:50 Kyle but wait and see if they oh want to
1:38:53 interview me okay I'll I'll do that
1:38:55 let's see what it
1:38:57 does Crystal's new and hasn't seen the
1:39:00 janky oh you will so it you won't see
1:39:04 the janky initially it's it's
1:39:07 fascinating especially if if you're good
1:39:10 if you're a good human Communicator like
1:39:13 if you're good at talking to people and
1:39:15 you just talk to the the AI whether it's
1:39:18 an image generation tool or chat GPT or
1:39:21 or you know notebook LM
1:39:24 and you just have like a good back and
1:39:26 forth conversation with it you will get
1:39:28 really good results it's when you want
1:39:31 to do something
1:39:34 specific where it'll get it like it'll
1:39:37 get it like 80 or 90% there where you're
1:39:40 like oh that's great oh this one
1:39:42 sentence and then you try to get it to
1:39:44 fix that sentence and it [ __ ]
1:39:46 everything else up and then and you're
1:39:48 like wait but well I want a part of this
1:39:50 and a part of that but now this is right
1:39:52 but that's wrong and it's just this
1:39:54 never ending back and forth and and what
1:39:56 you end up with is every single AI
1:39:59 project is essentially just a
1:40:01 Frankenstein monstrosity of piece parts
1:40:04 that it got some part of right and then
1:40:07 you know you're gluing them together
1:40:09 with your actual human
1:40:11 words it's [ __ ]
1:40:14 wild yeah like adversarial prompts are
1:40:17 really good yeah like no that sucks
1:40:19 don't do it that way I don't like that
1:40:20 make it less words make it less wordy
1:40:22 make it more wordy make it it flowery
1:40:24 don't make it flowery all that sort of
1:40:26 stuff's brilliant the more context it
1:40:29 has has the better it gets all right
1:40:31 here we go again Welcome to our deep
1:40:33 died into AI Festivus 2024 yeah it was a
1:40:36 two-day online ENT it really felt like a
1:40:39 global pep rally for AI enthusiasts and
1:40:43 curious minds alike it really was a
1:40:45 global Gathering people were tuning in
1:40:47 from all over Algeria Australia you name
1:40:50 it and what really stood out was the
1:40:52 focus on excessive ility making AI
1:40:55 understandable and approachable even if
1:40:57 you aren't a tech expert right it wasn't
1:40:59 about complex algorithm or coding jargon
1:41:02 it was about the human side of AI how
1:41:04 it's impacting Our Lives its ethical
1:41:06 implications and how we can all be a
1:41:08 part of shaping its future and speaking
1:41:10 of accessibility the event was
1:41:12 completely free which is a testament to
1:41:14 the organizer's commitment to making AI
1:41:17 knowledge inclusive absolutely it was a
1:41:19 huge success I mean the event drew a
1:41:21 massive online this is one of the things
1:41:23 that this uh feature will sometimes do
1:41:27 is what it's doing right now where
1:41:28 they're kind of each finish finishing
1:41:31 each other sentences but it's like it's
1:41:33 totally mechanical it's like every other
1:41:35 line it's it's very annoying this this
1:41:38 is one I would probably throw my
1:41:39 audience and there was a real Buzz on
1:41:42 social media during and after the event
1:41:44 with people sharing their excitement and
1:41:46 key takeaways absolutely the energy was
1:41:48 contagious and it's no surprise
1:41:50 considering the hosts an Murphy founder
1:41:53 of the shal lead's AI Community andal
1:41:55 Shannon the digital innovator behind the
1:41:57 AI Salon which boasts over 2,000 members
1:42:01 they really curated a unique experience
1:42:03 focus on positive applications of AI
1:42:05 speaking of community the event was
1:42:07 sponsored by a whole network of AI
1:42:09 focused organizations and podcasts
1:42:11 including localized podcast coaching
1:42:13 cubic squared marketing AI build lab and
1:42:17 empowered by AI podcast it really
1:42:19 showcased the collaborative Spirit of
1:42:21 the AI That's nice it pul it pulled the
1:42:23 sponsors out of the out of the
1:42:24 transcripts that's great the world it
1:42:27 did okay so let's dive into the sessions
1:42:29 right day one kicked off with scaling AI
1:42:32 from individuals to the Enterprise this
1:42:35 panel discussion featured Jessica reong
1:42:37 a prominent AI thought
1:42:40 leader um all right because I actually
1:42:43 want to use this I'm gonna go I'm Gonna
1:42:47 Change I'm going to redo it
1:42:50 copy all right let's see go delete
1:42:55 delete customize paste all right start
1:42:58 out with an overview of the
1:43:04 event that
1:43:07 includes how
1:43:09 popular it was and the
1:43:15 stats
1:43:17 the
1:43:19 attendance
1:43:22 stats the intention of the event and a
1:43:26 little
1:43:33 bit and their
1:43:37 communities the AI salon and she
1:43:44 leads
1:43:47 AI all right um
1:44:02 give and
1:44:07 overview uh
1:44:09 [Music]
1:44:42 there we
1:44:44 go gotta be less than 500 characters
1:44:47 damn it hang on uh make this less
1:44:54 and
1:44:56 500 characters I know this is boring I
1:45:00 know this is boring I've said this
1:45:03 before when you actually want to do
1:45:06 something that's usable it doesn't make
1:45:09 for
1:45:15 good that's not it okay damn it let's
1:45:20 create short summary of each session
1:45:25 highlights from their
1:45:27 session
1:45:33 and come on this has got to be less than
1:45:36 500 characters
1:45:55 what I don't get it I don't get
1:46:19 it oh refresh yeah maybe
1:46:23 that's a good
1:46:25 idea reload the
1:46:29 page
1:46:31 customize bang
1:46:36 nope
1:46:37 give
1:46:39 overview
1:46:46 of that includes
1:47:02 okay I just got I got to go back to chat
1:47:03 gbt uh make this 500 characters or less
1:47:10 and don't lose meaning
1:47:22 [Music]
1:47:28 provide an overview include attendant
1:47:30 stats purpose and
1:47:49 summarize generate okay good Lord good
1:47:53 lordy lordy lordy I was thinking these
1:47:56 two have gotten really cozy since their
1:47:59 last show something a
1:48:01 foot
1:48:03 uh if people watch me do AI work it
1:48:06 would be me cursing the numbers they
1:48:09 would die of boredom
1:48:12 Pate Pate so um Pate modder who's here
1:48:16 Dr Dr Pate modder uh which I realized
1:48:19 Pate on your session we didn't put Dr P
1:48:22 modder in front of your name so my
1:48:23 apologies for that because that's shitty
1:48:25 because he's a real doctor he's a PhD um
1:48:28 anyway um P did this really cool session
1:48:31 where he built us a a Google collab
1:48:34 notebook which if you don't know what
1:48:36 that is it's basically like a Jupiter
1:48:37 notebook it's think of it like he took
1:48:40 the time to write some code that does a
1:48:43 bunch of cool stuff and all you have to
1:48:44 do is hit play buttons but you can go in
1:48:45 and you can modify the code and he and
1:48:47 he created this whole thing where you
1:48:49 can go in and muck with the code and he
1:48:51 explains what the code does and it
1:48:53 explains how large language models work
1:48:55 and it actually visualized how a large
1:48:58 language model breaks um the English
1:49:02 language like a paragraph into tokens
1:49:05 into the individual piece parts that
1:49:07 make up words and punctuation and all
1:49:10 that sort of stuff and then it it shows
1:49:13 how it turns that into just mathematical
1:49:16 SE like numerical sequences and when
1:49:20 it's doing all of what it's doing it's
1:49:21 literally just it's like the Matrix
1:49:24 right you see all those numbers falling
1:49:26 it's just dealing with math man it's
1:49:29 just math up in the [ __ ] Laten space
1:49:33 doing its thing and then it comes out
1:49:35 like it comes out it's absolutely
1:49:38 remarkable oh man I don't care nobody
1:49:41 calls me doctor okay good I just I felt
1:49:44 bad after the fact because someone
1:49:45 pointed it out Serena pointed it out um
1:49:49 um Amelio's wife pointed it out and she
1:49:51 sent me a she sent me email that I
1:49:53 didn't see until after the event it said
1:49:56 important it was like all caps P modder
1:49:59 is a doctor so Serena head your
1:50:05 back
1:50:06 so what are you gonna do we we put the
1:50:09 whole thing together in 15 days
1:50:13 so all right we'll let this thing
1:50:15 generate and we'll see if it it got any
1:50:17 better if not then I'll just bail on it
1:50:20 but um hopefully you learned a little
1:50:22 bit about what notebook LM is it's
1:50:25 pretty remarkable oh did we oh let's
1:50:28 let's go back while it's generating that
1:50:30 thing let's go back here and go
1:50:32 um give me the
1:50:36 top 10
1:50:39 tips you
1:50:41 think the
1:50:43 participants
1:50:45 of this event would
1:50:49 give to
1:50:51 someone new to to
1:50:54 AI if they want to
1:50:58 be in quote marks AI
1:51:05 ready okay tomorrow at 11:00 a.m.
1:51:09 Mountain Time 10: a.m. Pacific 1 pm
1:51:12 Eastern an Murphy and I who co-hosted AI
1:51:15 festiv us we're going to be on um a live
1:51:19 stream I'm going to go on AI learning
1:51:21 lab so we'll have kind of the back scene
1:51:24 thing like we do here on on Tik Tok and
1:51:27 then it'll be live on LinkedIn and I'm
1:51:29 pretty sure YouTube um so if you go to
1:51:31 my LinkedIn Channel um and also
1:51:35 somewhere in the YouTube comments Vicky
1:51:37 posted the URL for it but my LinkedIn
1:51:40 channel is Kyle Shannon Kyle Shannon on
1:51:43 LinkedIn you can find me there um and
1:51:48 what Ann and I are putting together
1:51:49 we're going to launch it February 1 is
1:51:51 we're putting together an AI Readiness
1:51:54 certification based on AI Festivus what
1:51:57 we realized one was that people wanted
1:52:00 this they wanted like can this could
1:52:02 this be a certif certified thing I'm
1:52:04 like well it isn't like we just threw it
1:52:06 together in in two days but there is a
1:52:09 bunch of stuff there that we're going to
1:52:12 turn into a certification program so
1:52:15 we're going to do an AMA about that
1:52:17 tomorrow at 11:00 a.m Mountain 10
1:52:20 Pacific 1 Eastern
1:52:28 um tomorrow night I'll be back doing
1:52:31 regular AI learning lab
1:52:33 duties um if you're on YouTube right now
1:52:37 or if you could go to YouTube and
1:52:40 subscribe I'm trying to get to a
1:52:41 thousand subscribers the good news is I
1:52:43 cracked 700 the bad news is I haven't
1:52:46 cracked
1:52:47 800 so it's going slow so I need
1:52:50 subscribers on YouTube so the YouTube
1:52:52 channel is learning la- AI on YouTube
1:52:58 and go subscribe to that and then that
1:53:00 way if the Tik Tock thing goes away um
1:53:03 hopefully I'm at a thousand before Tik
1:53:05 Tok goes away it looks like they might
1:53:07 do some sort of last minute something or
1:53:09 other for
1:53:15 that um and then just mark on your
1:53:19 calendars a week from next T not next
1:53:24 Tuesday it's not the seventh is it it
1:53:28 is yes it is the seventh
1:53:33 okay yeah yeah yeah yeah next Tu uh next
1:53:38 Tuesday January
1:53:40 7th is um the next AI Salon meeting uh
1:53:46 we have a speaker I forget who the
1:53:47 speaker is I've got to go grab that info
1:53:49 um but you should come to the AI Salon
1:53:51 meeting think like AI T of us but much
1:53:53 shorter and once a
1:53:58 month all right uh so we got that let's
1:54:01 go see what the top 10 tips were so
1:54:03 based on AI Festivus here's the here are
1:54:06 the recommendations that notebook LM
1:54:09 came up with um based on what was
1:54:13 presented over two days if you want to
1:54:15 be AI ready here's the 10 most important
1:54:17 things you should focus on I'm actually
1:54:20 going to post this into the AI Salon
1:54:22 this is really good number one start
1:54:24 with a problem not a tool huge focus on
1:54:28 identifying specific pain points or
1:54:30 problems you're facing and then explore
1:54:33 how AI tools can solve them rather than
1:54:35 starting with the technology itself this
1:54:37 is why we say play first then mindfully
1:54:40 create this is talking about mindfully
1:54:42 creating start with a problem play just
1:54:45 go do anything with AI doesn't matter
1:54:49 then number one okay Embrace a
1:54:52 beginner's mindset approach AI with
1:54:54 curiosity and a willingness to
1:54:55 experiment rather than feeling
1:54:57 intimidated or overwhelmed it's okay not
1:54:59 to know everything and learning by doing
1:55:01 is a great way to get started just start
1:55:04 using it and brain dump don't hesitate
1:55:06 just jump into using AI tools just like
1:55:08 I did with the question about how can we
1:55:11 start a side business just start talking
1:55:12 to it you don't need to be structured or
1:55:14 organized just experience experiment and
1:55:17 see what happens AI can understand
1:55:19 unstructured input that is so important
1:55:23 AI can understand
1:55:25 unstructured input you can literally
1:55:28 just vomit into this thing and it will
1:55:31 figure it out it's
1:55:33 amazing focus on practical quick wins
1:55:35 look for small manageable tasks where a
1:55:38 AI can make an immediate impact rather
1:55:40 than trying to tackle the complexities
1:55:42 right away this approach allows you to
1:55:44 build confidence and demonstrate the
1:55:45 value of AI quickly use free AI tools to
1:55:48 get started begin with free tools like
1:55:50 chat GPT or Claud familiarize yourself
1:55:52 with their capabilities this lets you
1:55:54 learn and experiment the without the
1:55:56 pressure of financial commitments I also
1:55:58 would say don't buy any annual
1:56:01 subscriptions right now just do month-to
1:56:03 month leverage AI to enhance your zone
1:56:06 of Genius brilliant identify your
1:56:08 existing strengths and areas of
1:56:10 experience and explore how AI can help
1:56:12 Elevate them love that be aware of data
1:56:15 privacy and security understand the
1:56:16 importance of cyber hygiene and data
1:56:18 security when using AI avoid putting
1:56:20 sensitive information into public AI
1:56:22 tools redact personal and proprietary
1:56:25 information great love
1:56:26 it build your AI intuition focus on
1:56:30 understanding how to interact with AI
1:56:32 tools and prompt engineering start by
1:56:34 learning the fundamentals like
1:56:35 understanding the tools and how to ask
1:56:37 questions or give directions build
1:56:39 community love this join communities
1:56:42 like the AI Salon or she leads AI to
1:56:44 learn from others to share your
1:56:46 experiences collaboration and connection
1:56:48 can support can provide support and
1:56:50 insights focused on your adaptability
1:56:52 quotient and emotional quotient um Adnan
1:56:56 and uh Brian are going to be thrilled
1:56:58 with this one that's that's from their
1:56:59 session from their new book uh focus on
1:57:02 adaptability quoti and emotional quoti
1:57:05 as AI leads to Rapid change focus on
1:57:07 developing adaptability and emotional
1:57:09 intelligence in yourself and your teams
1:57:12 as well as technical skills [ __ ]
1:57:14 brilliant this is
1:57:16 great um so I'm going to
1:57:20 go share this over on a
1:57:23 s it's French
1:57:27 [Music]
1:57:30 [ __ ] yes I'm singing the Olympics
1:57:37 theme we'll go to AI
1:57:41 Festivus uh I'm G hide this so
1:57:45 notebook
1:57:47 LM filled with AI fested give
1:57:54 us
1:57:58 goodness just created what I feel is a
1:58:05 oh we're going to do this in Show and
1:58:08 Tell we'll go here we'll go here we'll
1:58:12 go here
1:58:18 uh hang on a
1:58:21 sec the copy and paste out of
1:58:25 uh out of these tools is really bad
1:58:32 unfortunately let's
1:58:40 see oops
1:59:05 [Music]
1:59:14 go here copy out of here go back over
1:59:17 here go here go here
1:59:20 H God damn it
1:59:26 it aches why does stuff suck so
1:59:30 bad
1:59:33 why why select all copy no see select
1:59:40 all
1:59:42 copy stuff shouldn't be so hard but it
1:59:47 is but it is
1:59:54 [Music]
1:59:58 oh my
2:00:05 [Laughter]
2:00:11 God oh my God
2:00:14 why
2:00:17 why do you know there's some engineer
2:00:20 somewhere probably that works with hate
2:00:23 that could [ __ ] fix
2:00:28 this
2:00:29 well what you cease to understand is
2:00:33 that there are what you're probably
2:00:35 using is a asy 24
2:00:38 73.2 uh protocol there and uh if if you
2:00:41 use the the the more well you could if
2:00:45 if you want to focus on backwards
2:00:47 compatibility you could focus not so
2:00:50 much on aski but ansy uh ancy well and
2:00:53 you've got the heximal there's a
2:00:55 situation with the the heximal focus of
2:00:58 the situation that is causing a this is
2:01:02 probably user error at this point simply
2:01:04 if you had gone into a a traditional
2:01:06 text editor which would have stripped
2:01:08 out the additional characters uh
2:01:10 wouldn't be such a
2:01:12 challenge good flipping
2:01:17 Lord um
2:01:20 notebook note
2:01:23 book
2:01:29 LM made
2:01:32 this AI
2:01:37 Readiness list and I
2:01:43 think it is
2:01:46 brilliant notebook LM um
2:01:53 full of AI festiv us
2:01:58 goodies made this Ai and I think it's
2:02:03 brilliant and then we're going to say
2:02:06 select all that copy it and then we're
2:02:08 GNA go down here and we're going to post
2:02:11 it so we're gonna not notify everyone
2:02:15 because that's [ __ ] obnoxious and
2:02:17 then we're going to go here we're going
2:02:19 to post it here as well because I think
2:02:20 it's good in both
2:02:32 places
2:02:34 wow all right there you go there you
2:02:38 have it
2:02:40 people all right and put ourselves out
2:02:42 of a job we got to keep things semi
2:02:44 broken exactly
2:02:48 listen everyone that says computers
2:02:51 programmers are not going to be needed
2:02:52 they're going to be needed as long as
2:02:54 you can't open a [ __ ] webp image
2:02:57 format you remember the show Syble
2:03:00 Kyle's the poster
2:03:06 child all right so that's done all right
2:03:10 let's let's get out of here let's get
2:03:11 the [ __ ] out of here people what am I
2:03:12 doing with with your lives um happy New
2:03:15 Year um come back here tomorrow come
2:03:18 hang out tomorrow at 11:00 a.m. mountain
2:03:21 time 10:00 a.m. Pacific Pacific 1 pm.
2:03:23 Eastern and learn about the new AI
2:03:26 Readiness uh certification that we are
2:03:29 putting together we're working with uh
2:03:31 with uh Vicky and an and I are
2:03:36 gonna smash that thing into some
2:03:39 semblance of order um and it should be
2:03:41 really good it it was actually a really
2:03:43 good experiment tonight working with
2:03:45 notebook LM because it's the you know
2:03:48 it's these tools are just staggeringly
2:03:50 good all right I think I've covered
2:03:53 everything um I'm going to get out of
2:03:55 here thank you all so much thank you for
2:03:58 the entertainment and knowledge the
2:04:00 entertainment little weak knowledge thin
2:04:04 thin thin and thin thin and
2:04:07 thinner except for
2:04:09 me tubby
2:04:11 mington all right peace out everyone
2:04:13 I'll see you later