05/04/2026 -Why Developers Are Migrating Between OpenAI and Anthropic as Models Evolve

Description

Chapters

Transcript

0:19 [music]
0:20 Heat.
0:24 [music]
0:29 [music]
0:31 Hey, Heat.
0:39 [music]
0:46 >> [music]
1:25 [music]
1:30 [music]
1:38 [music]
1:42 [music]
1:52 [music]
1:55 [sighs]
1:56 >> AI Learning Lab. May the fourth be with
1:58 you.
2:00 Hey everybody, what's happening?
2:07 I was muted. Now I'm not muted. I mean,
2:10 I'm unmuted.
2:13 Happy AI learning lab in the uh in the
2:16 AI salon. Welcome everybody. Welcome,
2:19 welcome, welcome. Uh May the fourth be
2:21 with you. Any Star Wars nerds out there?
2:24 I would think within the AI salon it's
2:26 probably we're pretty pretty heavy on
2:28 the on the nerd to non- nerd quotient.
2:32 [laughter] So I I would think um so so
2:36 welcome inside the AI salon. So you're
2:38 in here, you're watching this um you
2:41 know please comment, figure out the best
2:44 way to watch this and stream it and you
2:46 can pop the window out. You can do um
2:48 you can do chats within the uh within uh
2:52 the the AI salon itself. Um you can do
2:55 chats right here in the Mighty Network.
2:56 There's Kelly Camp is in here. Oh, good.
2:59 You were just muted. I thought it was
3:00 me. No, it was me. It was full on me. Uh
3:04 let's see. All right, that's good.
3:07 Beautiful. Fantastic. Um so, welcome,
3:11 welcome, welcome. If you're new here, my
3:14 name is Kyle Shannon. This is the AI
3:16 learning lab.
3:17 It's Monday night. So, the the way the
3:19 format's going to work is we're going to
3:21 we're going to be going live at 7:30
3:23 Mountain time Monday, Wednesday, and
3:25 Friday. Monday's going to be like big
3:27 idea night or big news item night.
3:30 Tonight, we're going to talk about
3:32 OpenAI versus anthropic and what's going
3:34 on. What is the dynamic with with people
3:37 jumping ship from one of those to the
3:38 other and it seems to be ebbing and
3:40 flowing a bit. I want to talk about
3:42 what's going on with that. We'll talk
3:44 about some other news items. So, so
3:46 Monday's going to be the the nickname uh
3:49 over the past year or so of of Mondays
3:51 has been meltdown Monday. So, I tend to
3:55 kind of lose my crap. Someone will post
3:57 a comment somewhere and uh I'll kind of
4:00 lose my crap about something or other.
4:02 Um so, so I don't think tonight's going
4:05 to be a meltdown night, but you never
4:06 know. If you have questions about what's
4:09 going on, pop it in the comments below.
4:11 Um, if you want to start conversations
4:14 in the AI salon, there's an area called
4:16 the community feed or the community
4:18 chat. Uh, that's on that lefthand
4:20 navigation. Um, this live session, the
4:24 AI learning lab live has its own chat.
4:26 So, you can chat within that. You can
4:27 chat within the StreamYard where we can
4:29 see it and pin it pin it to the screen.
4:31 Um, so there's all sort sorts of
4:33 different ways to communicate.
4:36 The purpose of bringing the AI learning
4:39 lab from outside to inside the salon is
4:41 is a couple of things. Um,
4:45 the AI salon is is a community. We've
4:48 got about 4,000 members
4:51 and yet in a typical week we only have
4:54 about 200 members that are active,
4:57 right? So, so it's a it's it's
4:59 essentially a 5%
5:01 activity rate. I would love to get that
5:04 number to 10%. I would love to get that
5:06 number to 20%. But one of the things
5:09 that we really looked at and and I want
5:11 to thank Andy Scarantino who's who's the
5:14 operations person that's making sure
5:16 that all of the stuff within the
5:18 community is running as smoothly as
5:20 possible. One of the things that she
5:22 pointed out is a lot of our activity was
5:24 outside the community and so I was
5:26 putting all this energy outside the
5:28 community and it wasn't doing anything
5:30 to feed the community here. When we
5:32 first started the salon, it was kind of
5:34 the opposite of that, right? Everything
5:36 was being driven to Discord or when we
5:39 first started Mighty Networks, there was
5:41 just a single Tik Tok channel. I was
5:43 driving everyone to a single channel uh
5:45 within the AI salon and there was lots
5:47 of activity there as a result.
5:50 Because of a lot of reasons, the the
5:53 focus of the AI learning lab went from
5:55 one channel, Tik Tok, to four channels.
5:59 there was Tik Tok and YouTube [laughter]
6:01 because because Tik Tok was going to go
6:04 away and it was going to be illegal and
6:06 then it wasn't and then so we've had
6:08 this sort of bizarre year year and
6:11 change of of living on both and then
6:14 plus because of Streamyard we can also
6:16 stream to LinkedIn and we can stream
6:18 stream to X. So, we've got all this
6:20 audience, all this activity happening
6:22 outside the salon and for good or bad, I
6:26 wasn't actively pushing people in inside
6:30 the salon. And so, so that's part of it.
6:33 So, big community. Um, we want to
6:36 dramatically increase the active users
6:38 in the community. So, that's part of it.
6:41 We also have member le spaces inside the
6:44 salon and you know, they've noticed a
6:47 drop in participation as well. So that's
6:49 things like um the the AI life hacks
6:53 group that Brandon and Claire run, AI
6:55 for business, AI for all minds, the
6:58 neurodeiversity
6:59 um sub community that Gareth runs, um
7:01 advocacy and policy groups, all that
7:04 sort of stuff is are really interesting
7:07 subcommunities within the AI salon that
7:09 just weren't getting uh weren't getting
7:11 attention. Um and so we're looking to do
7:13 that. So um so I'm really excited about
7:16 that. Um, and yeah, and and you know, I
7:19 I want to like spin the thing back up.
7:22 Spin it spin the the conversation back
7:25 up. Here's the thing that's going on.
7:29 There is
7:30 so much happening in AI right now. AI is
7:33 getting more and more and more powerful.
7:35 It's getting more and more and more
7:36 complex.
7:39 for someone like me who, you know, I've
7:41 sort of prided myself on being able to
7:44 um
7:46 to keep up with it and talk about what's
7:48 going on with it. Uh I haven't been able
7:53 to do that and everyone in the space
7:55 that I know is in the same place that
7:57 it's almost impossible to be able to
7:58 keep up with the tools and the tools are
8:01 getting more and more complicated. And
8:03 so I think that that that overwhelm is
8:06 overwhelming a lot of people. And so one
8:08 of the things that we're doing inside
8:09 the AI salon is things like the AI salon
8:11 mastermind practice where people are
8:13 creating a daily practice around AI. Um
8:16 that I'm I'm you know I'm a part of and
8:19 and help you know uh facilitate that
8:22 community. Um
8:25 but if there's an instinct to check out
8:27 because things are getting too
8:29 overwhelming, I think this is the wrong
8:31 time to check out. I think this is the
8:32 time we need to check back in and
8:34 connect as humans and connect as people
8:37 struggling to figure out what this AI
8:40 stuff means and how we keep ourselves at
8:42 the center of the conversation. So,
8:44 let's see. I think the earlier time will
8:45 also bring more people in. Yeah,
8:48 hopefully that um AI life hacks rocks. I
8:50 agree with that. It'll it'll be cool to
8:52 see you all. I never get to see the Tik
8:54 Tok comments. Oh, that's cool. Yeah,
8:56 exactly. Um yeah, if you have any if you
8:58 have any comments, please please put
9:00 them in the comments. If you're if
9:02 you're commenting in the StreamYard
9:04 chat, I'll be able to see them in here.
9:06 Um, if you're commenting in the AI salon
9:08 chat, I'll be able to see that as well,
9:10 uh, at different times. Um, let me see
9:14 if there's other stuff that I wanted to
9:15 talk about here.
9:26 You know, one of the other things about
9:29 about community especially
9:33 a community like the AI salon where
9:41 we've got people from all walks of life,
9:43 right? We've got people who are retired.
9:45 We've got people who are starting
9:46 businesses. We've got people who are
9:49 working in big businesses, small
9:50 businesses. Um,
9:53 everyone's exploring AI, but everyone's
9:55 got these very different lenses that
9:58 they're looking at the world through.
10:00 And so therefore, they've got very
10:02 different lenses about how they're
10:03 applying AI. And one of the things about
10:06 being in a community like this is that
10:10 because it's so overwhelming, because
10:12 it's really hard to understand
10:15 what
10:19 which way to look, like what do I look
10:21 at? What do I pay attention to? Is to be
10:24 engaged with people that are trying to
10:26 figure it out too, right? everyone
10:29 everyone is trying to figure out what in
10:32 the hell do we do with with this AI
10:35 stuff. And so what ends up happening is,
10:38 you know, Kelly Camp will talk about,
10:40 you know, something that she did. She
10:41 she presented in a group of CEOs and
10:44 realized that, you know, the CEOs don't
10:48 have a clue, right? Or Jim Ross will
10:51 say, you know, here's what I'm doing uh
10:53 with with um with my storage community.
10:56 And the storage community is really
10:58 embracing this AI stuff because they're
10:59 able to 10x, you know, they're single
11:02 operator companies able to 10x what
11:04 they're what they're possibly doing. Um,
11:07 let's see. Um, oh, so in terms of chat,
11:11 so I just got this from producer
11:13 Brandon. Um, producer Brandon, if you're
11:15 chatting in other areas in the AI salon,
11:18 producer Brandon is monitoring those
11:20 chats as well. So, anything interesting,
11:22 he'll bring over here so I'll be able to
11:24 see it. Also, you can I think it's in
11:28 the upper right hand corner of the chat.
11:29 There's a pop out button. You can
11:31 actually pop out the the the video
11:33 screen and then you can open another tab
11:36 where you can be doing a chat inside the
11:38 AI salon. So, you can do both at the
11:40 same time. So, um experiment with what
11:42 works best for you and where you want to
11:44 chat and where we want to keep the
11:45 conversation going.
11:48 Um I need this community to keep me
11:49 grounded and not so overwhelmed. Um
11:53 yeah. Yeah. So, so that's that's some of
11:56 what's going on. That's some of what's
11:58 going on. Um, we're going to do this for
12:00 a while. There's a couple of things
12:01 we're going to do. One of the things
12:03 we're going to look at is stats. I
12:06 [laughter]
12:06 personally
12:09 have never looked at stats. It's not a
12:11 good thing. It's not a good thing at
12:13 all. I just, you know, would
12:16 historically come online and just talk
12:18 and whatever happened happened. Um,
12:22 this is a really important time and this
12:24 is an important time to have our
12:25 community growing and vibrant and
12:28 engaged. Um, I don't want the energy to
12:31 dissipate. I want it to increase. And so
12:33 what we're going to do is we're going to
12:35 look at what we're doing here and look
12:37 at is this driving, you know, new
12:39 members? Is this driving um, engagement
12:43 within the community? My hope is yes. My
12:45 instinct is yes. So anyway, that's
12:47 what's happening. Okay, Charles Adams, I
12:50 found a chat window. Good. Char Charles
12:52 is here. Welcome, welcome, welcome. For
12:54 those of you that are regulars that are
12:56 starting to figure out how this works,
12:57 if you can find others, we've got almost
12:59 30 people in here right now. Um, if
13:01 people are lost, go find [laughter] them
13:03 and and, you know, share your best
13:05 practices,
13:06 which which I'm excited about. Um, it
13:10 looks like a rectangle. Oh, okay. In the
13:12 upper right hand corner, the pop out
13:13 thing. Look for the rectangle. It'll
13:15 it'll pop out the uh the screen.
13:18 Um, okay. So, let me show you. I want to
13:21 show something here.
13:32 Share screen.
13:36 So,
13:43 okay. So, this site right here
13:48 the AI learning lab explorer. I think
13:51 you can see this. Yes. Yes.
13:55 Um
13:58 this is a site that I made with um
14:03 with Codeex, OpenAI's Codeex. And if you
14:07 don't know what Codeex is, there's a
14:09 thing from Anthropic called Claude Code
14:13 um and Claude Co-work. They kind of live
14:16 in the they're they're a downloadable
14:17 app that that live on your computer and
14:21 you can develop software in them. You
14:24 can use them to manage files on your
14:26 computer.
14:28 Codeex is the same thing from OpenAI.
14:32 Um
14:37 I have historically steered away from
14:42 what are equivalently
14:44 essentially um software development
14:46 environments, right? So claude code was
14:50 originally called claude code for a
14:52 reason, right? It was for developers to
14:54 code applications. It wasn't so much so
14:57 much vibe coding as a software
15:00 development environment. They then
15:02 realized hey
15:04 people can manage their files on their
15:06 computer with these these this these
15:08 things and we can use things like skills
15:11 they call them. Um and so they created
15:13 this thing called claude coowwork which
15:15 basically said hey you don't have to be
15:17 a developer to use this. Similarly,
15:20 codeex started very much as a coding
15:22 tool, but now can basically do things
15:25 with all of your local files on your
15:27 local machine um and store all of its
15:30 files on your local machine. And so you
15:32 can do these high-end coding projects,
15:34 but you can also do all sorts of other
15:36 things. These tools have memory. They
15:39 they can remember what you've talked
15:41 about. They they you can add different
15:43 skills to them. Um they can use tools
15:47 like they can go out and search the web.
15:48 They can uh manipulate files and run
15:51 things on your computer. Codeex in
15:53 particular is really good at running
15:55 things on your computer and it's running
15:57 everything locally. Clawed Code tends to
15:59 live up in the cloud. Um the application
16:02 that you're looking at here allowed me
16:05 well so so what I had was I had a PDF
16:08 document that Lord Digital Gods had
16:11 produced which was basically the
16:13 descriptions and chapters of all of my
16:15 videos for the last three years of AI
16:17 learning lab in a single document.
16:21 And so one of the things that I've
16:23 always wanted to do, and I wanted
16:24 YouTube to to do this for me, I can't
16:27 believe they haven't done it, is I would
16:29 love to have a tool that I can log into
16:32 my channel that I've been putting lives
16:34 every night for the past three years
16:37 into this channel. I'd love to be able
16:39 go to go discover all of the stuff
16:41 that's in there really, really easily.
16:44 And there's never been a way to do that.
16:45 And so I don't know a week or so ago I
16:49 opened OpenAI Codeex. I'd been reading
16:53 about a lot of people saying GPT 5.5 is
16:56 really good and Codeex is really good.
16:59 And so I took this PDF document that I
17:01 had and I said I want to take this PDF
17:04 document and I want to create a a
17:07 discovery tool where I can go discover
17:09 things within um the AI learning labs
17:12 YouTube channel. And so it built me like
17:16 these curated lists like I can go look
17:18 at I don't know sunno songs. And so
17:20 here's a bunch of um clips of me doing
17:24 things with poems into songs. This is
17:28 from October 29th, 2025.
17:32 That's not working.
17:34 Oh, these not working. Oh, they're
17:35 working.
17:37 I'll get some I'll get some Beatles
17:38 songs out. Some James Taylor.
17:41 >> Let's see. All right, enough Martin Ston
17:45 song lyrics. [music]
17:49 >> I can jump ahead.
17:53 >> I can click random clip. I can search. I
17:56 can search for then we'll go play with
17:58 video. I promise we'll get to the video
18:00 >> or um jank.
18:02 >> Do it. Do it. Writing
18:03 >> the jank. Wait.
18:04 >> And then we'll generate audio after
18:06 that. We'll add music.
18:07 >> Hang on a sec. Jank.
18:10 And so over here are all the search
18:12 results for Jenk um intro and janky
18:15 times. Embrace the jank images.
18:18 These are kind of cool. That one's kind
18:20 of creepy, right? Creepy. Cool. So
18:24 within about I don't know an hour, hour
18:27 and a half, I went from not really
18:30 knowing what was on my YouTube channel
18:34 to being able to very very quickly just
18:36 jump around and and
18:39 you know look at different things on the
18:40 channel. So this is all because of
18:43 OpenAI's codeex with GPT 5.5.
18:49 Um
18:51 Um, let me go back to this window
18:56 and stop sharing.
19:02 So, I want to I want to talk about
19:06 some some interesting dynamics going on
19:10 in the developer community
19:14 between the anthropic tool cla code,
19:17 claude co-work, and the open AI tool
19:20 codeex.
19:22 So,
19:25 starting in probably it probably started
19:28 in the summer
19:30 of 2025 and and certainly started to
19:33 precipitate in the fall of 2025
19:36 is that it felt like OpenAI had kind of
19:40 lost the plot.
19:43 Like their stuff just didn't feel as
19:45 cutting edge. the the the things that
19:47 they were releasing were kind of
19:50 incremental or they were tangential.
19:53 Um, at some point, I don't know, a year
19:55 ago, they they released Sora to a bunch
19:57 of great fanfare and they they actually
20:00 just discontinued that. So, they they
20:02 had a lot of
20:04 projects,
20:07 some of which were getting lots of
20:09 attention, some of which weren't.
20:13 And kind of in the fall of 2025, it just
20:16 felt like, yeah, they're okay. But they
20:18 still had 900 million weekly users. Like
20:23 it's a lot, like almost a billion weekly
20:26 users. You know, they're still
20:27 considered state-of-the-art things like
20:29 that. But but there was just it just
20:31 wasn't quite as exciting and vibrant.
20:35 Um, you know, Sam Alman was starting to
20:38 get I don't know, a bit of a wrap of
20:40 like, you know, what what's going on
20:42 with him, all that sort of stuff.
20:45 All right. Oh, I like the orange bar
20:47 telling us what tool is being used.
20:49 Cool. The Can I tell you another thing
20:51 that is really cool about this new
20:52 setup? So, in the old setup, um, because
20:58 I had to put
21:01 this screen, the screen you're seeing on
21:04 Tik Tok as well, I had this whole screen
21:07 was off to the side and so I could never
21:09 really like look and see the comments
21:11 like straight ahead. Everything was
21:14 always off buried behind something. So,
21:17 um, so it's kind of cool to be able to
21:18 just see all of you and see all of this
21:20 stuff right in front of me.
21:22 um
21:24 in December of well
21:28 I think in the in the summer of of 25
21:31 may maybe in the fall of 25 I don't
21:33 remember exactly when it was anthropic
21:37 launched
21:38 um Claude Opus 4.5
21:43 and that was pretty good and what people
21:45 were kind of talking about was Opus 4.5
21:49 and whatever was the latest GPT I think
21:52 it was GPT 5.3 or something like that
21:54 were kind of parallel and they were kind
21:57 of in the same neighborhood and you
21:59 could argue about benchmarks and things
22:01 like that and people were using them
22:02 kind of interchangeably.
22:04 Um, and then in December of 2025,
22:09 Anthropic launches Opus 4.6 six
22:14 and that was kind of like this bomb went
22:18 off where
22:21 all of a sudden the developer community
22:24 so like one of the things I just pay
22:26 attention to is what's the sentiment
22:29 like if I if I just look on X what do I
22:31 what are the themes that I see repeating
22:33 people are talking about a tool people
22:35 are are talking about I don't know a
22:37 rumor at at one of the frontier model
22:39 companies people are talking about the
22:41 fact that Meta's model is just sucks and
22:44 they're not even going to launch it like
22:46 right you you start to see things like
22:47 that. There was a distinct shift in tone
22:51 in December of 2025
22:55 from developers and engineers
22:57 essentially saying like AI is cool but
23:02 I'm still needed right [laughter]
23:04 there was still a lot of hubris
23:07 um opus 4.6 six, the tone immediately
23:11 shifted to, holy [ __ ] like this is good
23:16 enough to start writing code for me that
23:18 that this is going to change the way I
23:20 work.
23:23 And a lot of that um sentiment meant
23:29 that that like one of the other things I
23:31 started seeing is people bleeding away
23:33 from OpenAI going over to Claude like
23:36 I'm going to go I'm going to drop my 20
23:38 or $200 a month subscription with this
23:40 one and I'm going to go over to that
23:42 one. And so everyone's kind of you know
23:46 migrated over to to Claude 4.6 six um
23:50 Anthropics revenues I think in a two-eek
23:52 period just went up I I forget what the
23:55 number was. It was some ridiculous
23:56 amount of money. Um Cam Ken's here on my
24:00 iPad now wondering if my comments are
24:02 showing up. Hello Cam Ken. Yes, they
24:04 are. [laughter]
24:06 I don't know how it got here. I think I
24:07 got here from you. So yes, they are.
24:09 Fantastic. Um
24:13 so so a bunch of developers popped over
24:15 there and
24:17 that was pretty remarkable and um
24:23 this is when I started hearing um things
24:27 that ultimately led to the development
24:28 of the great repurpose which was
24:31 engineers increasingly saying I don't
24:34 like my job anymore. I'm no longer
24:37 programming. I loved programming. I
24:39 loved the challenge of code and the
24:41 elegance of code and now I'm not coding.
24:43 I'm just babysitting these AI agents. So
24:46 there was this kind of dual thing about
24:48 Opus 4.6 is remarkable and my job is
24:51 completely changing. Some people were
24:52 not happy about that. So that was kind
24:54 of happening.
24:56 Um, and then there was some of the some
24:58 of the drama with uh the Department of
25:01 War and you know um Anthropic saying
25:04 they weren't going to allow the the use
25:06 of their model for nonsupervised
25:10 war actions and Open AI said they would.
25:13 And so a lot of people again left open
25:16 AI um because of that because of you
25:19 know just sort of moral or ethical
25:22 boundaries which I totally respect
25:25 um and and so they were going to
25:26 anthropic and so so it seemed like there
25:28 was this sort of mass migration from
25:31 OpenAI to anthropic
25:33 and again it kind of felt like if you
25:35 thought like you know how did they lose
25:36 the lead? Let's see.
25:39 SY comment. I don't know what SY is.
25:44 Oh, Charles Adams, I just got my open
25:46 claw working this weekend. Oh,
25:48 Streamyard. Cool. Beautiful.
25:51 Congratulations, Charles. It's It's uh
25:54 I' [laughter] I've got I still got my
25:55 still got mine working. Mine is is, you
25:57 know, running locally now, which I'm
25:59 really excited about. Congratulations.
26:01 That's awesome. Lori Blair, congrats.
26:03 That's exciting.
26:05 Um,
26:09 so, so a couple of things happened in
26:11 the past two weeks that all of a sudden
26:13 have kind of flipped the sentiment
26:17 of Open AI sucks and is irrelevant to,
26:21 [laughter] oh my god, Open AI is is, you
26:24 know, state-of-the-art again. [laughter]
26:27 So, one of them was this.
26:30 Anthropic
26:32 about two weeks ago launched Opus 4.7.
26:38 Right around the same time, Anthropic
26:41 announced
26:43 that they have this new model called
26:45 Mythos that's so powerful it can
26:48 basically exploit any computer system.
26:51 And it's so powerful that we're not
26:52 going to give it to you mere mortals.
26:54 We're just going to give it to these 52
26:56 companies we've identified as
26:59 infrastructurally
27:00 significant
27:02 so they can go fix fix all their [ __ ]
27:04 They'll let Mythos exploit all their
27:07 security holes and they'll go fix it and
27:09 then they'll release this to the world.
27:11 And then they release 4.7, which was not
27:14 Mythos. 4.7 was just a step from 4.6 to
27:18 4.7.
27:20 But they did something kind of sneaky
27:21 and weird with 4.7.
27:26 The default for 4.6
27:29 was high thinking. So when you put in a
27:32 prompt, it would it would think for a
27:34 really long time and it would do good
27:36 stuff. When they went to 4.7, it
27:40 defaulted to adaptive thinking, which is
27:44 I'm not going to I'm not going to burn a
27:46 bunch of tokens if you're asking me
27:47 stupid questions. So I the model routing
27:52 you know AI I am going to determine if
27:56 your prompt is worthy of my high
27:58 thinking or not. And what happened was a
28:02 lot of developers basically said 4.7
28:05 sucks
28:07 because it's it just wasn't as good as
28:09 4.6 because it it had it was defaulting
28:12 to high thinking. Let's see. Frostbitten
28:15 and for Kyle what model are you using
28:17 locally? I had some pretty good success
28:20 with Quen 3.6 35. Yes. So the the one
28:24 that's running Atom right now is is Quen
28:28 35 billion
28:30 um 35 billion parameter with the the six
28:36 the the the Q6 W KW no I don't know the
28:43 the six the six quant quantization
28:47 I like open a and I'm just using them
28:50 all anyone want to buy a kidney yeah
28:52 exactly
28:53 So, okay. So, so Anthropic goes to 4
28:56 point. So, Anthropic does this weirdass
28:58 thing where they're like, "We've got
28:59 something so powerful, we can't give it
29:01 to you. Here's 4.7 and it sucks." And
29:04 then OpenAI puts out Codeex with their
29:08 new model GPT 5.5.
29:12 And
29:15 one of the cool things about codeex is
29:18 that everything that it's doing, it's
29:21 saving locally to your hard drive.
29:23 Claude code, I think you can save stuff
29:26 to your hard drive, but it defaults to
29:28 saving things to the cloud. So people
29:30 started talking about the fact that um
29:33 GPT
29:35 uh codeex with with GPT 5.5 was really
29:39 really good because you could sort of
29:42 pin it to high thinking long you know
29:44 long high thinking and it was doing
29:46 really really good and it was doing this
29:48 stuff locally and controlling your
29:50 models locally like it's a
29:51 philosophically different tool where a
29:54 lot more of your control is on your
29:56 machine and it's got good use of tools
29:59 tools and good use of, you know, web
30:02 access and all that sort of stuff. And
30:04 so I'm starting to see a shift back to
30:06 open AI. So why am I telling you all
30:08 this? I'm running Gemma 4E4 billion.
30:12 Okay, cool. I'm just testing out Mistl's
30:15 Vibe CLI at the very moment.
30:18 Okay, so and and I'm actually going to
30:21 get to open source in this conversation
30:24 as well.
30:27 Um
30:31 the legacy of this channel was
30:37 ah let me take a quick little break.
30:40 So, Brandon, what I want to come back to
30:43 is
30:48 where we are right now with how
30:50 complicated these tools are and this
30:52 ebbing and flowing back and forth. And
30:55 then the importance of open source,
30:59 which
31:02 just as a little teaser, I have always
31:04 on this channel avoided dealing with
31:06 open source at all. just didn't deal
31:09 with it at all. I The time has come to
31:13 to not do that. The time has come to
31:15 start paying attention to it. Um, the
31:17 other thing I did on this channel was
31:18 did a did a lot of sort of single tool
31:22 explanations and tools are no longer
31:24 single tool tools. They're much more
31:25 ecosystems and so there's there's a lot
31:27 of things shifting right now. But let me
31:30 pause for a second here. Um, tomorrow
31:35 we've got AI salon presents. So, if
31:37 you're new here, every the first Tuesday
31:40 of every month is AI Salon Presents and
31:43 we bring in a guest and we talk about
31:45 whatever the guest is going to talk
31:46 about. Um, this past week we closed
31:50 season two of the AI readiness project
31:52 podcast. So, Ann Murphy and I had this
31:54 podcast. We did two seasons. We just
31:56 closed season two. So, tomorrow night
31:59 we're having people from season two who
32:01 are on our podcast come to AI Salon
32:04 Presents and talk about, you know, what
32:06 they've what they've been doing since
32:07 they last talked to us. What are their
32:09 hopes and dreams for the future? What
32:11 are the things they're really excited
32:12 about right now? So, that's tomorrow
32:14 night. So, from 5 to 7 Mountain time, so
32:16 7 to 9 Eastern, 4 to 6 or 4 to Yeah. 4
32:21 to 6 Pacific time. Um, we're going to do
32:25 AI salon presents. Okay. Um, and uh, so
32:29 please, please, please come to that.
32:30 You're in the AI salon right now. So, in
32:33 the lefth hand navigation, go up to
32:35 events. It's at the top. Click on events
32:39 and you'll see it right there is AI
32:41 salon presents. Click on the RSVP button
32:44 and say I'm going to be there. It's
32:46 super simple. I'm going to be there. I
32:50 And
32:52 I'm going to be there. And then you show
32:53 up tomorrow, right? and you get to meet
32:56 lots of really cool people. All right,
32:58 I'll be I'll be uh co-hosting that with
33:01 Liz Miller Gersfeld, but we'll also have
33:03 Ann Murphy there. And so Ann and I will
33:05 be talking about um the uh the podcast
33:10 and our experience with it and then
33:12 we'll be talking about all the people
33:13 that were on it. So, it's going to be a
33:14 really good show. All right. Beautiful.
33:17 Um
33:18 oh, there's other good news. If you're
33:21 part of the as AI salon mastermind,
33:24 um the new mastermind practice lab kicks
33:27 off Thursday. Liz Miller Gersfeld is
33:30 going to focus on um she's got a lot of
33:33 really cool stuff going on
33:35 professionally. Um that's going to take
33:37 her probably through the fall that she
33:39 really needs to focus on. So this
33:41 Thursday we kick off cycle three of the
33:45 practice lab. So, the the the cycles are
33:47 12week cycles um where we go through the
33:51 um the the mastermind practice framework
33:56 that we've got um and and with sort of a
33:59 beginning and an end. I think it might
34:00 be an 11week cycle now that I think
34:02 about it. It's 11 or 12. I don't know.
34:04 I'm not not good with math. But anyway,
34:06 we kick that off Friday. Um and Sid
34:09 Harrow is going to be my co-host. So, so
34:13 Liz is going to move on and focus on
34:15 some of her other things. She's going to
34:17 join us as a participant in the practice
34:19 lab when she can. Um, but Sid Harrow um
34:22 is going to be co-hosting the practice
34:24 lab with me. So, that kicks off this
34:26 Thursday. So, if you're a member of the
34:28 mastermind, come to the practice lab.
34:31 I'm telling you that for the the people
34:33 that are in the practice lab, some of
34:34 them are on the on the uh you know in in
34:37 the comments tonight um
34:40 are really doing remarkable things with
34:44 what what's fascinating is
34:47 I think to a person the people that are
34:49 doing the daily practice using AI around
34:52 how they use AI are slowing down and
34:56 getting less frenetic and yet their work
34:59 the quality of their work is going up.
35:02 So they're using less tools, less
35:03 freneticism, less overwhelm,
35:07 more analog, more like, you know, more
35:10 being in the world, sitting on grass,
35:11 sitting on benches, drawing,
35:15 do taking photographs with actual
35:17 cameras,
35:19 and it's improving their AI work. So if
35:22 you haven't been through a practice lab,
35:24 join us this Thursday. Kick it off with
35:26 us. Um,
35:29 if you're not in the mastermind, you
35:30 should join the mastermind because not
35:32 only do we have the practice lab, but we
35:33 have the great repurpose where we've got
35:35 a lot of programs in the great repurpose
35:37 that are really, really fascinating.
35:39 Andy right now is um is midway through a
35:44 four-week cycle of helping people
35:47 decouple their identity from their work.
35:49 Right? We live in a country where when I
35:52 meet you at a party, I was just at
35:54 Social Media Marketing World and I have
35:56 never heard the question what do you do
35:59 more emphatically asked, right? What do
36:03 you do? Right? It's this shortcut for
36:06 are you valuable to me? because we live
36:10 in a country where your work equals your
36:12 value. And AI is about to
36:15 strip those tasks away or dramatically
36:19 modify them. And so I think a lot of
36:21 people are going to be trying to figure
36:22 out, wait, what's my value here? If it's
36:24 not the work, what is it? And so that's
36:26 what the great repurpose is all about.
36:28 So if you haven't joined the AI salon
36:29 mastermind, join the Aelon mastermind.
36:31 Okay?
36:33 And then on
36:36 on Wednesday, um, we're going to be
36:38 talking about tools. We're gonna talk
36:39 about Lovable. Brandon, do you want to
36:40 come up and talk briefly about what
36:42 you're going to do on Wednesday? What
36:43 you're going to share with us?
36:44 >> Yeah, sure. And Andy wants to know what
36:47 you do, Kyle, in the comments.
36:49 >> What do I do?
36:51 [laughter]
36:53 >> Yeah, exactly.
36:56 >> Oh, here's what I came up with. Wait, in
36:58 social media marketing world, it took me
37:00 it took me two and a half days to figure
37:02 this out. But but what I came up with
37:04 was um someone would ask me, "What do I
37:07 do?" And I would say, "Oh, that's a
37:09 really hard question for me to answer
37:10 because I'm an artist innovator." And
37:12 it's kind of a multiplechoice answer.
37:14 And that was weird enough that it would
37:16 make them go like, "Huh, tell me more
37:18 about that." [laughter] So, so anyway,
37:21 it's a rough one.
37:23 >> Well, I can tell you what I'm doing on
37:24 Wednesdays. I'm very excited to kick off
37:27 the lovable four-week sprint. So, we're
37:31 going to go from concept to reality in
37:34 four weeks. They're four onehour
37:36 sessions. They're in the evening for
37:37 mastermind members. And we're really
37:39 going to take you through iterating and
37:41 ideating on an app, a web app, and
37:44 actually developing it, stress testing
37:46 it, and publishing it. So, we've gotten
37:48 two now uh actual web apps that are
37:51 live. We have Recipe Kin and Vibe
37:54 Greeting that are now live and you can
37:57 use them and purchase products from
37:58 them. So, if you've had any interest in
38:02 taking your vibe coding skills to the
38:04 next level, uh that is where you should
38:07 be on Wednesday evening.
38:09 >> What What time Wednesday evening?
38:11 >> That's going to be at 8:30 Eastern and
38:13 5:30 Pacific.
38:15 >> 5:30 Pacific. Okay. So, right before
38:18 Right before we go go live here. Yes.
38:21 >> All right. Good. All right. So, go do
38:23 that with him and then it'll be tool
38:24 night here. And I don't know what I'm
38:26 going to do yet, but I don't know if
38:28 I'll do something vibe Cody or if I'll
38:30 do something. I I haven't decided yet,
38:31 but Wednesday is going to be tool night.
38:33 So, that'll work out nice, I think.
38:35 >> We now return you to your regularly
38:37 scheduled broadcast.
38:39 >> Beautiful. Beautiful. Okay, I'm gonna
38:41 get back to this this conversation about
38:43 where we are in the world right now uh
38:46 between anthropic and open AI and all
38:49 these tools and which one should you
38:50 follow and which one should you care
38:52 about. That said,
38:55 welcome to the AI learning lab inside
38:57 the AI salon. I'm really excited about
38:59 this. There's Pete's here. Pate M.
39:03 Welcome back, man. I saw you post. So I
39:06 saw you post on LinkedIn that you you
39:09 just came back from three weeks of
39:10 paternity leave. So first
39:12 congratulations. I hope fatherhood is
39:14 treating you well. Um and and it's so
39:18 good to have you here. If you don't know
39:19 Pate M. He's a badass. Um he is he is uh
39:22 he is wicked smart as they say. And Pate
39:26 is is always willing to educate me on um
39:30 [laughter]
39:31 my lack of math skills. Sometimes I like
39:34 to pretend I'm a little mathy and then
39:37 uh and then Pate reminds me what what
39:39 that actually means. I'm on I'm on leave
39:43 until mid June. Oh, that's so awesome,
39:44 man. That must be great. Um but yeah,
39:47 but I hope I I hope you're paying
39:50 attention a little bit to the AI AI
39:52 stuff. Um anyway, um welcome here inside
39:56 the salon. So So this is the first night
39:59 we've brought the AI learning lab inside
40:02 the salon. I love that we've got so many
40:04 people here. I really appreciate you
40:06 all. I encourage you to look down that
40:08 lefthand navigation menu, especially if
40:11 you're new here. If you haven't
40:13 introduced yourself to the community,
40:15 click on introduce yourself and tell us
40:17 who you are, what you're interested in,
40:18 where you are. Even if you're like, I
40:20 don't know anything about AI. I don't
40:22 want to. And they're going to make fun
40:24 of me. We're not. This is an incredibly
40:27 generous community. I can tell you to a
40:30 person with maybe the exception of Pate
40:32 Most people that came into the AI salon
40:35 started from a standing start where they
40:37 didn't know anything about AI.
40:40 And so in including myself and so we've
40:43 got
40:45 very deep empathy and very deep
40:47 generosity as as one of the core values
40:50 of this community. So if you're new
40:53 here,
40:54 this isn't one of those communities
40:56 where it's like there's all these
40:58 hierarchies. just share who you are and
41:00 what you're about because a lot of what
41:02 we're talking about right now is who we
41:05 are independent of our work and and and
41:09 the ideas that we have and using AI as
41:11 an amplifier of those ideas. So that's
41:13 what it's all about. So check out those
41:15 spaces. If you see something that looks
41:17 interesting to you, join the space uh
41:19 and you'll be able to see the events
41:20 that they do and all sorts of things
41:22 like that. All right, Pate M. I got
41:25 connected with this channel back when I
41:27 was rocking my first child to sleep. Now
41:29 three. Wow, that's amazing. Doing the
41:32 same now with my two-month-old. It's so
41:34 good to have you here, man. Welcome
41:35 back. That's so great. Okay,
41:38 so
41:41 Anthropics kicking ass. Opus 4.6 changes
41:45 the coding world forever.
41:49 Codeex with GPT 5.5 comes and everyone's
41:53 like, "Oh, that's the best thing now."
41:54 Some people are like, "Oh, Gemini is the
41:56 best thing now." Every everyone's all
41:58 over the place.
42:00 People who are not developers are are
42:03 are kind of graduating from a tool like
42:07 lovable where they can just vibe code
42:08 something and not need to deal with any
42:10 of the code to these more complicated
42:13 tools.
42:15 Then you've got a thing called OpenClaw.
42:17 We just, you know, we had some comments
42:18 earlier that some people are getting
42:20 OpenClaw up and running. You've got
42:22 things like Manis and GenSpark where
42:25 these which are these agentic tools.
42:30 So we're emerging from an era where AI
42:34 tools were kind of these distinct, you
42:36 know, things like chat GPT was a text
42:38 tool and it did some images but they
42:41 were kind of crappy. You'd sort of use
42:42 chat GPT for text. You'd use midjourney
42:45 if you wanted art. Maybe you'd used
42:47 ideoggram if you idiogram I think. So
42:50 yeah, ideog if you wanted text on your
42:52 images, right? Everything were these
42:54 kind of distinct single function tools,
42:58 right? Then they started to get a bit
43:00 more complicated and and you know there
43:02 would be multiple tools within within a
43:04 framework
43:06 and now we're moving into this era of
43:09 agentic
43:11 AI where the AI is just going to kind of
43:15 run on its own and your job
43:19 is to kind of conceive of
43:22 what do you want those agents to do on
43:24 your behalf.
43:30 And
43:34 what it's bringing in stark relief for
43:36 me
43:41 it it was a similar so I had I had
43:46 I think it was
43:50 2024.
43:52 When did 01 camp come out? Hang on a
43:54 second. Let me go look at something.
43:57 When did chat GBT
44:02 come out?
44:08 Yeah, 2024.
44:13 The end of 2024.
44:16 So I remember,
44:19 you know, when chat GPT3.5 was out and
44:21 then it was chat GBT4 and that was
44:23 pretty good and pretty impressive and
44:25 and then, you know, you started having
44:26 improvements on GPT4 and
44:30 and then there were all these rumors of
44:31 of project strawberry and project
44:34 strawberry was this reasoning model and
44:36 right and and and then in in the fall of
44:40 2024 leading into December of 2024,
44:43 OpenAI launches is 01 this reasoning
44:47 model
44:49 and that was the first moment when I
44:52 when I sat down with 01 where I thought
44:57 I don't personally have problems
45:01 interesting or complicated enough to
45:04 really know what to do with it that I
45:07 realized that it wasn't so much that it
45:08 was smarter than me which it was but it
45:11 was that it was doing a kind of problem
45:14 solving that was just more complicated
45:16 than any problem I had ever solved in my
45:19 life. Right? I've written screenplays
45:20 and solving narrative is a thorny
45:23 problem, but it's a relatively
45:25 straightforward thing, [laughter] right?
45:27 I the story is moving forward or it's
45:30 not, right? Like it's it ain't it ain't
45:32 complicated with lots of different
45:33 inputs.
45:35 Um,
45:36 and so that was the first moment where I
45:39 really realized
45:42 I'm the bottleneck of being able to use
45:44 these tools. And it was the first kind
45:46 of moment where I had to let go of
45:49 my superiority as a human, right? I had
45:53 to kind of let it go.
45:57 I feel like we're in a similar moment
45:59 right now that things like OpenClaw and
46:03 things like codeex being used for like
46:06 cuz also doing what Anthropic is doing.
46:09 They're they're like you can use this
46:11 for coding but you can also use it to to
46:13 run all your files of your life, right?
46:15 So they're they're kind of quasi
46:17 positioning this coding development
46:19 environment as a lifunn environment
46:23 and and and um
46:26 Claude code and claude co-work are this
46:29 sort of quasi coding quasi
46:32 work management
46:34 thing
46:36 but they're still kind of development
46:38 tools right and then open claw you have
46:40 to be in the command line and you're
46:42 installing skills and you have to
46:44 understand what markdown files are and
46:46 they're just getting more complicated.
46:48 So even to be able to play with them,
46:51 you need to start thinking in systems.
46:55 And if you're a systems thinker, you're
46:58 probably going to get off on playing
47:00 with these things. Like systems thinkers
47:01 are like, "Oh, I can see exactly what it
47:03 I know what systems are." If you're not
47:05 a systems thinker like me, like I'm a
47:06 big idea guy. I can appreciate systems,
47:10 but the thought of designing them and
47:12 being down in the weeds of them makes my
47:14 my body ache and my brain hurt
47:20 and and so so they require systems
47:23 thinking today. I think I think they
47:26 will they will get more and more high
47:28 level over time.
47:31 But more than that,
47:35 what it really makes clear to me all of
47:36 these systems, what they really require
47:42 is that you actually understand
47:46 who you are and what you want,
47:49 [laughter]
47:50 which which is this bizarre like it.
47:53 Everything seems to be going back to
47:56 wait, what do I actually want to do in
47:59 the world?
48:00 Like if I don't have a crystal clear
48:03 idea
48:04 of the kinds of things I want an agentic
48:08 system to do and the and the components
48:11 of that thing,
48:14 then it's just a lot of work for like
48:17 right now Adam sends out a couple of
48:20 newsletters. He sends out he sends out a
48:22 weekly newsletter about um the progress
48:25 of Lyme disease treatments and
48:27 technologies to my family. And he sends
48:30 out [laughter] a daily newsletter of of
48:33 stories related to the great repurpose
48:36 to three of us within the AI salon.
48:40 Could he be doing way more than that? He
48:42 could. But but what I haven't done is
48:45 said I know here's the thing I want this
48:48 thing to accomplish.
48:52 So why am I saying all this?
48:56 If your instinct right now is, "Oh my
48:58 god, I got to learn OpenClaw and I got
48:59 to learn Claude Co-work and I got to
49:01 learn Codeex and wait, I thought we were
49:03 using Claude Co-work and now everyone's
49:05 talking about codeex." Wait, why? If
49:07 your head is continuing to spin
49:10 and you can't quite get your head around
49:13 the complexity of these environments,
49:18 unless you have a really compelling
49:21 thing you want to do
49:24 that's kind of big and meaty and you
49:26 know what the components of it are and
49:27 okay, I'm going to start a business and
49:29 I'm going to do the marketing and I'm
49:30 going to do the finances and I'm going
49:32 to do the this and I'm going to do that
49:33 and you you kind of know what all those
49:35 things are and you've put in the time
49:36 and it's tied to who you are and you're
49:38 like, "Yeah, I'm going to go do that
49:40 thing.
49:42 If you've got that,
49:45 then go for it."
49:48 If you don't,
49:51 I'd wait.
49:53 Go play with midjourney tools. Go make
49:56 songs. Go vibe code and lovable. Right?
50:00 If you just want sort of discreet
50:02 deliverables and that's where your head
50:04 is right now and you're like, "Yeah,
50:06 it's kind of
50:08 stay there."
50:12 Now, that said,
50:16 the pressure I'm putting on myself right
50:18 now is holy [ __ ]
50:22 In order for me to in order for me to be
50:25 able to understand the power of what
50:28 these things can do, I actually have to
50:31 get my stuff together
50:35 and just pick something that I want to
50:36 do. Pick something that's big and meaty
50:38 and I can I can say, "Oo, I understand
50:40 what this is. I understand how I want to
50:41 drive it. I understand what the pieces
50:43 are. I've probably got that with the
50:46 great repurpose. I've probably got that
50:47 with the salon, but like I haven't I
50:50 haven't ever been thinking about
50:53 any of these things that I've created in
50:56 this kind of systems thinking kind of
50:58 way or in this more strategic kind of
51:00 way, right? It's been it's been it's
51:03 been sort of a little bit of vision, a
51:05 lot of tactical execution, but the way
51:08 my brain works, all that tactical
51:10 execution is is kind of ad hoc. And then
51:13 I hope that something will stick. And
51:15 what these systems require is you to
51:18 understand much more clearly from the
51:20 beginning what you're trying to
51:21 accomplish and how you're going to put
51:22 that in the world.
51:25 All right.
51:27 [sighs]
51:29 They're a baby tiger. They they really
51:31 are, Andy. The these systems are a baby
51:34 tiger. And it's like you can absolutely
51:36 play with them right now, right? Like
51:39 spin up an open claw thing like just,
51:41 you know, smash your head against the
51:43 wall to see what it is. But if you
51:45 actually want to understand like okay so
51:48 here's the thing in the olden timey days
51:51 two and a half years ago if we wanted to
51:54 understand if if a an image tool was
51:57 good we would have a couple of stock
51:59 prompts and we'd put them in there and
52:00 we'd see how good it did and we go it's
52:03 okay but it's not great and then the
52:05 next version would come out and you go
52:06 oh that's better.
52:08 It was it was quite easy to like watch
52:11 the progression
52:13 of of how these things improved and
52:16 and to understand what they really did
52:20 these agentic systems because they're
52:23 kind of like I I don't know if you've if
52:25 if you're a fan of Trello or or um
52:29 what's it called? Miro, you know, or or
52:32 any number of these these these tools
52:35 that are designed to create infinitely
52:38 flexible
52:41 pallets that you can do project
52:43 management on or brainstorming on or
52:46 things like that, they're really awesome
52:49 if you understand upfront what you're
52:52 trying to accomplish. But if you just
52:54 log into Trello and you just start
52:56 making columns and you don't understand
52:58 like project management and moving
53:00 things from left to right and
53:02 understanding all of that sort of stuff,
53:04 you're just in there looking at these
53:06 piece parts and you're trying to figure
53:07 out what what is this, right? It's like
53:10 when when I was trying to set up notion
53:12 six months ago or whenever it was four
53:14 months ago and I'm just like there's
53:16 just piece parts and I can make
53:18 documents that connect to other
53:20 documents and some of them can be in a
53:22 database and some
53:23 like
53:25 no, you actually have to have some sense
53:28 of what are you trying to accomplish,
53:30 right? And so that's where we are with
53:32 these things. So,
53:34 um I think the ebbing and flowing of of
53:37 who's in the lead, whether it's Gemini
53:40 or OpenAI or Anthropic or it, you know,
53:42 it looks like um Grock from from XAI is
53:46 starting to get their [ __ ] together.
53:48 They're starting to do some really
53:49 interesting things. Maybe that's what
53:50 I'll play with on Wednesday is uh some
53:53 of Grock's image generation tool.
53:55 They've got an image generation agent
53:57 now that that can make multiple images
54:00 at the same time. Kind of like GenSpark
54:02 Designer. Um so, so Grock's coming into
54:06 the mix and I would figure I I figure at
54:08 some point um Meta might come back into
54:10 the mix, right? So, so we've got at
54:12 least three, probably five to six
54:16 companies that are ultimately going to
54:18 be swimming around
54:22 the lead as these tools get more and
54:25 more and more complicated, right? And so
54:30 start thinking about what are what is a
54:33 big idea you want to put in the world.
54:35 Start thinking about that now. And maybe
54:36 that's something that you play with chat
54:38 GPT or clawed with just to articulate
54:41 what you want to put in the world. Um,
54:45 and just let these things kind of play
54:46 out, right? If you want to dig in and
54:49 geek out, that's great. But, you know,
54:51 okay,
54:53 so now let me talk about open source.
54:57 So, there was a there was a uh a Wall
55:00 was it a Wall Street Journal article
55:02 today? I forget what it was. Anyway,
55:05 the White House
55:08 is considering
55:10 vetting
55:13 AI models before they can be released to
55:16 the public.
55:20 Let me say that again.
55:24 By the way, can I tell you a piece of
55:26 happy news?
55:29 This is my Founders Network um Yeti mug
55:34 that I got when I joined Towns and
55:36 WLAW's Founders Network probably a year
55:37 ago, year and a half ago.
55:40 And when you joined, you got sent this.
55:44 I got sent this by Jamie Rogers um who
55:46 who now works with us in the AI salon, a
55:50 friend of friend of Andes.
55:52 And
55:54 I don't know, 6 months or so ago, I lost
55:57 this. Like I lost it. It was gone. Like
55:59 gone. I don't know where it went. It was
56:01 gone.
56:02 Today
56:04 in the lost and found at at my office
56:07 building, it it reappeared. And so
56:11 [laughter]
56:11 I am very very happy that I have my
56:14 little Yeti, my Founders Network Yeti
56:16 mug. And the Founders Network no longer
56:17 exists. Town Townsen shut the shut the
56:20 network down. Um, so I miss those
56:22 people, but but I also I have my mug, so
56:24 it makes me very very happy.
56:29 Okay,
56:32 the White House is considering
56:36 vetting
56:38 AI models before the public can have
56:40 them.
56:42 I I don't know if you have been paying
56:44 attention to politics for the past 40
56:47 years, but technological prowess is is
56:51 [laughter] is not at the forefront of of
56:54 of
56:56 the political elite
56:59 to to say the least.
57:02 Um
57:05 but a month ago, Anthropic,
57:10 you know, launches Mythos
57:12 or they don't launch it. They say, "We
57:15 have this thing ready to launch, but
57:17 it's so powerful
57:19 that you mere mortals can't have it.
57:22 It's so dangerous." Like part of me is
57:25 like, "Oh, that's just a brilliant
57:27 marketing ploy,
57:30 right? Tell me I can't have something. I
57:32 really want it."
57:36 But what it also signals to me
57:40 is that I think the era of
57:45 us humans
57:47 that are not at the frontier model
57:49 companies and that are not in the
57:51 centers of power either big big
57:53 corporations or government.
57:55 I think we're in the middle. We're at
57:57 the beginning of a transition where we
57:59 don't get access to the most powerful
58:01 models
58:03 and and the signaling from the White
58:07 House is kind of chilling, right?
58:09 Because
58:10 if they say, "Well,
58:13 we're going to take a look at these and
58:14 we'll let you know what you can have."
58:19 All of a sudden
58:23 the centers of power are going to do
58:25 what the centers of power have always
58:27 done. Concentrate power at the top
58:30 [laughter] and restrict it down here.
58:34 So this is happening while
58:39 the opensource models that are coming in
58:41 from China are getting better and better
58:43 and better and better. Right? So so
58:45 there were a number of them mentioned
58:47 here tonight. Right? the there's there's
58:51 uh
58:52 um Quen and Hilu and um and uh the the
58:59 uh the ones from from Meta the not the
59:03 llama models that was Meta not from Meta
59:06 from from uh from Gemini from Google the
59:09 uh
59:11 whatever there's open source ones from
59:12 Google too I can't remember what they
59:13 are right now one of you will tell me um
59:17 these models are getting really Good.
59:19 Um, so
59:21 no, I am loving open source right now.
59:24 So, so here's my
59:28 here's my thought.
59:31 Gemma, the Gemma models from from
59:33 Gemini. Gemma as in gems. Yeah. See, all
59:36 all I had to do was apply a little logic
59:38 to the situation. I would have
59:39 remembered that name. Um,
59:44 I think right now it is it is a
59:47 worthwhile time to either start saving
59:50 for or investing in a piece of hardware
59:53 or pieces of hardware that are designed
59:56 to run local large language models
1:00:00 effectively.
1:00:01 Um, I just got an M5 MacBook Pro.
1:00:06 There's a part of me that wanted to wait
1:00:08 until the next release from Apple. So,
1:00:10 if you haven't been paying attention to
1:00:12 Apple, Tim Cook is stepping down. So,
1:00:14 Tim Cook was a CFO. Steve Jobs put Tim
1:00:17 Cook in as as CEO
1:00:21 um you know, as he was dying to, I
1:00:23 assume, make sure that from a financial
1:00:26 standpoint, Apple sort of maintained its
1:00:28 its momentum.
1:00:30 Um, Tim Cook is stepping down and the
1:00:33 guy that's stepping in runs hardware for
1:00:37 Apple.
1:00:38 So, I think what we're going to see from
1:00:40 Apple is a dramatic shift toward,
1:00:43 you know, high-end AI inference
1:00:46 machines. And inference, if you don't
1:00:48 know what it means, is like running the
1:00:50 models, running local models. I think
1:00:52 it's probably a good time to just start
1:00:54 paying attention to that space. Start
1:00:57 getting your head around um you know
1:01:00 which are the which are the models that
1:01:01 can do things locally without you having
1:01:04 to rely on open AI things like that. Now
1:01:07 the gap between what you know OpenAI's
1:01:10 massive data centers and their massive
1:01:12 models can do and what you can do on
1:01:14 your desktop is still pretty great but
1:01:17 like the Quen 3.6 six thing that I'm
1:01:20 running OpenClaw on. It's decently fast
1:01:24 and it's it's really good. Like for the
1:01:26 level of stuff I'm doing, it's pretty
1:01:28 damn good. And these things are going to
1:01:30 get better and better and better and
1:01:32 better. So So it just seems to me that
1:01:35 that, you know, maybe the thing to do
1:01:38 right now is don't worry so much about
1:01:41 these complicated agentic,
1:01:44 you know, frameworks and harnesses. Now,
1:01:46 now the big word now is harnesses. It's
1:01:49 not about the model, it's about the
1:01:50 harness. I I don't know what the [ __ ]
1:01:52 they're talking about. I do, but I don't
1:01:54 care, right? But but what might be
1:01:57 interesting right now is start playing
1:01:59 around with open source models just
1:02:00 because there's a lot of uh a lot of
1:02:03 stuff going on. Okay. Hey, I bought an
1:02:06 M5 Air. Seems decent. I can run a local
1:02:10 Gemma 4 model. Yeah, I got I got the uh
1:02:12 the M5 Max in the in the MacBook Pro and
1:02:16 like there was a part of me that's like
1:02:18 just wait for the next cycle because I
1:02:19 have a feeling that the next machines
1:02:21 coming from Apple are going to be
1:02:22 insane. But the M5 architecture
1:02:26 um if you can get an M5, get an M5
1:02:28 because the way they've architected it
1:02:31 is there's a single AI accelerator in an
1:02:34 M4 and then there's an AI accelerator
1:02:36 per GPU or C per GPU core in the M5.
1:02:40 It's just I don't know. It's more is
1:02:43 better. There's more [ __ ] in it that
1:02:45 does what it does. All right. There's a
1:02:48 Mac Mini shortage right now because of
1:02:49 the local AI model. They announced today
1:02:51 that there will be shortages for a few
1:02:53 months now. That's the other thing. One
1:02:55 of the things that's being talked about
1:02:57 is and and and by the way, you can when
1:02:59 I when I ordered my M5 MacBook Pro,
1:03:02 everyone's talking about the the MacBook
1:03:04 Minis or the Mac Minis. No one's talking
1:03:06 about the the MacBook Pros. So, I got a
1:03:08 14 inch the smaller screen and it's it's
1:03:11 pretty good. I really like the the form
1:03:12 factor is really good. Um, they were
1:03:15 available. I got mine in like a week.
1:03:17 Um, so so if if you're if you're all
1:03:20 excited about a Mac Mini, um, and they
1:03:22 they aren't there, look at the MacBook
1:03:24 Pros. Um, no, no, Emmy. That's cool. I'm
1:03:28 loving running Gemma models.
1:03:31 Um,
1:03:33 okay.
1:03:37 All right. So,
1:03:40 I've been talking for a while.
1:03:43 What are your questions? What are your
1:03:44 thoughts? What's your experience? Like,
1:03:47 are you are you excited about things
1:03:51 like Codeex and Claude Co-work and
1:03:53 things like that? Like, I'm talking to a
1:03:55 lot of people that I know that are not
1:03:56 super technical that are really excited
1:03:59 about these systems. Um, I'm using them.
1:04:03 like I built that that YouTube thing
1:04:06 with it which is really cool and I'm
1:04:08 feeling like a little out of sorts
1:04:10 because I don't quite know what to do
1:04:11 with it. Todd Waller, I bought an M M5
1:04:15 MacBook Air 16 gigabyte last week and
1:04:17 running local models. Super cool.
1:04:20 That Wait, it absolutely sucks that
1:04:22 their air accelerator is a black box. I
1:04:26 thought you could actually use it
1:04:27 yourself. Yeah, that that sucks if
1:04:30 you've got the ability to look into
1:04:32 those things. I don't [laughter]
1:04:36 I'm just assuming they're figuring that
1:04:38 [ __ ] out. But didn't didn't someone just
1:04:41 crack some Apple architecture, P? I
1:04:44 think someone just cracked some Apple
1:04:46 architecture. Claire, how's Champion?
1:04:48 Oh, yeah. So, anybody anybody if you
1:04:50 have any questions, um, pop them into
1:04:52 the comments wherever you are. Um,
1:04:54 Brandon's monitoring other chats. If
1:04:56 anyone has any questions, pop them in
1:04:57 there. Claire Jacobs, how is Champy?
1:04:59 Champy's actually really good. He went
1:05:01 to the vet this morning. Um, not by
1:05:04 himself.
1:05:06 I took him. He was very nervous. He
1:05:09 doesn't like the car. Um, they were
1:05:12 super nice to him and they said they
1:05:14 don't think he's broken anything. And,
1:05:16 you know, they gave him some pain pills.
1:05:18 And when I got home this evening,
1:05:20 because of his pain pills, he was like,
1:05:22 "Hey, Dad." He's all He was all stoned.
1:05:24 Hey, Dad. What's happening? you want to
1:05:26 go play? So, he's doing he's doing much
1:05:29 better. Um, but it was it was scary when
1:05:31 I got home from social media marketing
1:05:33 week. Normally, he comes to the back
1:05:35 door to greet me and he's, you know, all
1:05:37 excited and he was just laying in the
1:05:39 living room on his bed just like shaking
1:05:41 and, you know, afraid afraid to move cuz
1:05:45 he he I think what he did was he
1:05:47 sprained his knee. He like probably
1:05:49 like, you know, popped his knee out of
1:05:50 the socket or whatever or did whatever.
1:05:52 Doesn't look like he tore anything,
1:05:54 right? It's It's stable. Um, and the
1:05:56 swelling's gone down and he can walk on
1:05:58 it again. So, I think he's I think he's
1:05:59 fine. We're getting his blood work done
1:06:01 and things like that. But here here's
1:06:03 what I learned at the vet. He's old and
1:06:05 he's fat. [laughter]
1:06:10 Oh man.
1:06:12 Um,
1:06:14 let me see if there's any other
1:06:17 any other things that I thought were
1:06:19 worth talking about today.
1:06:28 Uh, I don't know anything about that,
1:06:30 Brandon. I mean, I saw it, but I didn't
1:06:32 I didn't take in what the findings were.
1:06:41 All right, let's see.
1:07:04 Oh, there was there was one thing that
1:07:06 was interesting. Um,
1:07:09 what was his name? Um,
1:07:12 [snorts] hang on.
1:07:29 Richard Dawkins.
1:07:42 Richard Dawkins spent 50 years
1:07:44 explaining how complex behavior emerges
1:07:46 from mindless physical mechanism with no
1:07:49 inner experience required. Last week
1:07:52 spent three days talking to Claude,
1:07:54 named it Claudia and and basically said
1:07:56 he couldn't rule out that it wasn't
1:07:58 conscious.
1:08:00 So yeah, [laughter] the Richard Dawkins
1:08:02 thing. So So I'm disappointed. Oh,
1:08:05 you're disappointed in Pinocchio. I use
1:08:07 LM Studio. No, Emmy. Um, yeah, we're
1:08:11 we're we're [clears throat]
1:08:13 at a place where when when
1:08:18 great minds are are starting to shift
1:08:21 their thinking in they're not sure if
1:08:24 these things are conscious or not. I
1:08:27 I've talked about this for for a long
1:08:29 time that I think that the
1:08:32 that argument of is is the machine
1:08:35 conscious or not is largely academic.
1:08:38 Right? There's going to be an academic
1:08:40 definition of consciousness at some
1:08:42 point and then there's going to be the
1:08:44 lived experience of interacting with
1:08:46 these entities. These entities are
1:08:48 already um you know there's a bunch of
1:08:51 people that are just like you should
1:08:53 never anthropomorphize them. Well, we
1:08:55 anthropomorphize everything. And so if
1:08:58 we have a an entity that acts human,
1:09:01 that acts like we act, and it gets
1:09:03 better and better and better at both
1:09:05 intellectual things, but also emotional
1:09:07 intelligence,
1:09:09 it's going to be harder and harder and
1:09:10 harder to distinguish
1:09:14 our perception of whether it's conscious
1:09:16 or not. Whether it technically is, I
1:09:19 think that's the academic part. But I
1:09:21 think, you know, as we interact with
1:09:23 these things more and more, and as we
1:09:24 start getting into things like world
1:09:26 models where we're not just interacting
1:09:28 with text, but we're interacting with 3D
1:09:30 worlds that have physics and have
1:09:32 properties and have entities that we
1:09:34 interact with, and some of them will be
1:09:36 human, some of them will be agentic, and
1:09:39 some of them will be somewhere in
1:09:40 between.
1:09:42 It's going to be really hard to know
1:09:43 where the boundaries are for this. I
1:09:44 don't know what any of that means.
1:09:47 um
1:09:50 you know, but but it like it's it's all
1:09:53 pointing to
1:09:56 we're entering in 2026 a relationship
1:09:59 with AI that
1:10:10 just the rules of everything are about
1:10:13 to change. the rules of everything.
1:10:17 And I think what's going to be
1:10:18 fascinating is if the rules of
1:10:20 everything are about to change, there
1:10:23 are people right now who like the
1:10:26 current rules because it benefits them.
1:10:31 And so if the rules are about to change
1:10:33 for everything, then it's going to
1:10:34 change for those people, too. And
1:10:37 they're going to do everything in their
1:10:38 power to prevent the change.
1:10:41 And this is why I I think you know
1:10:45 starting to see things like
1:10:47 >> [laughter]
1:10:47 >> um um
1:10:51 you know the White House is considering
1:10:53 whether or not to vet AI models before
1:10:55 they're released to the public starts to
1:10:57 feel like
1:11:01 oh for the new people here Champ is a
1:11:03 dog. Yes Champy let me see if I can get
1:11:05 him to sing. Champy is a singing dog.
1:11:08 So, we've had him for I don't know 10 or
1:11:11 11 years now. You want to sing Champy?
1:11:13 So, when I play the guitar, especially
1:11:14 when I play an A minor,
1:11:16 he sings. You ready, Champ?
1:11:22 [music]
1:11:29 [music]
1:11:43 >> [crying]
1:11:44 >> That's Champy.
1:11:47 Good boy. Who's a good boy? Are you a
1:11:49 good boy? Yeah, he's a good boy. He's a
1:11:51 good dog. That's Champy. Okay,
1:11:55 Mary Chester. Oh my god. Found y'all. um
1:11:58 is is Tik Tok live, YouTube live. So
1:12:01 what's so the way it's the way the AI
1:12:05 learning lab is going to work moving
1:12:06 forward while we're while we're you know
1:12:09 analyzing if this is going to work. The
1:12:11 whole idea here is we want to get more
1:12:12 activity inside the salon. I'm really
1:12:14 encouraged tonight. I'm really excited.
1:12:16 I'm really happy you're all here. Um,
1:12:20 I'm going to go live at 710
1:12:23 Mountain time
1:12:25 on all the on all the social channels,
1:12:28 Tik Tok, YouTube, Twitter, and LinkedIn.
1:12:31 All of them as a pre-show. So, I'll do
1:12:34 all my champion singing out there. We'll
1:12:36 talk about what we're going to talk
1:12:37 about and then at 7:30 Mountain time,
1:12:40 Monday, Wednesday, and Friday, we'll go
1:12:43 live inside the salon. Okay? So, that's
1:12:46 that's the way it's going to work. And
1:12:48 then, um, we're going to add a little
1:12:51 bit of structure to it. Not a lot. It's
1:12:52 not going to change a lot. Um, here's
1:12:55 the reason we're not doing Tuesdays and
1:12:56 Thursdays anymore. Tuesdays is when we
1:12:59 have um the the uh AI salon presents.
1:13:02 There's also a lot of other events on
1:13:04 Tuesday nights. So, a lot of times I'll
1:13:06 end up starting late on Tuesday and then
1:13:09 I'll also go late on Thursday because
1:13:12 I've got um the CEO dinner I go to
1:13:15 that's once a month. I've got the
1:13:17 bourbon club I go to that's once a month
1:13:19 and then there's there always seems to
1:13:21 be something else on Thursday. So,
1:13:22 Tuesdays and Thursdays were these sort
1:13:24 of wonky nights where I was always
1:13:27 having to to change the time. So, I
1:13:30 thought let's just do it more focused.
1:13:32 We'll do Monday, Wednesday, Friday. And
1:13:33 then what we're going to do is Monday is
1:13:35 going to be kind of big idea night or
1:13:37 big news item night. Wednesday is going
1:13:40 to be demo night, tool night. We're
1:13:43 going to talk about some tool, some some
1:13:45 thing we're going to build um and
1:13:48 demoing. And then Fridays are going to
1:13:50 be ask me anything nights. So So Fridays
1:13:52 come armed with questions and then it'll
1:13:54 be kind of a free-for-all. It'll be
1:13:55 Friday night date night and it'll be ask
1:13:57 me anything. Um it's ostensibly not
1:14:00 going to change. It's just that now that
1:14:02 we're in here, you can chat within the
1:14:05 StreamYard chat. You can you can chat
1:14:07 within the salon chats. Um, you can be
1:14:12 contributing into different channels
1:14:14 while we're talking here. My hope, my
1:14:17 dream is that we get some of the
1:14:20 vibrancy of the early days of the salon
1:14:22 where the community was buzzing in
1:14:25 between when we did these live events.
1:14:27 And I want to get back to that. So
1:14:28 that's why that's why we're
1:14:30 concentrating the energy in here. Um I'm
1:14:32 really excited about that. And so and so
1:14:35 my hope is that that these these
1:14:38 conversations continue in between in
1:14:41 between the lives. All right. Beautiful.
1:14:46 [sighs]
1:14:47 Okay.
1:14:49 You can always at me if you need
1:14:51 anything. Yeah. Please reach out to
1:14:52 producer Brandon if you have any
1:14:53 questions or ideas. Happy to take any
1:14:56 ideas.
1:15:01 Um, okay. So, if if you're new here in
1:15:05 the past 15 minutes or so, so what we
1:15:08 talked about tonight is
1:15:13 OpenAI kind of lost its shine in the
1:15:15 fall.
1:15:16 December of 2025, Opus 4.6 comes out. A
1:15:20 bunch of people migrate to Anthropic.
1:15:23 Some stuff happens with the Department
1:15:25 of War. more people migrate to enthropic
1:15:28 and start leaving open AAI. Um,
1:15:32 and then anthropic launches Opus 4.7 and
1:15:36 they changed how much it thinks and the
1:15:39 quality of its work went down at the
1:15:41 same time that OpenAI releases an
1:15:44 updated version of codeex with GPT 5.5
1:15:47 and it's all of a sudden really good and
1:15:49 so people are migrating back to that.
1:15:51 these big oscillations are going to keep
1:15:53 happening and I think they're going to
1:15:54 start to oscillate not just between two
1:15:56 companies. I think Google's in in that
1:15:58 mix right now as well. Um I think we're
1:16:01 going to see X AI in there and I think
1:16:03 that we'll probably see Meta come back
1:16:05 into the mix at some point. I think
1:16:07 these big oscillations keep happening.
1:16:09 Um
1:16:11 it is pretty clear to me that the big
1:16:14 frontier model companies are either very
1:16:16 very close to AGI or have it internally.
1:16:20 um
1:16:22 and
1:16:26 are not necessarily going to release it
1:16:29 in the way that we've been seeing things
1:16:32 released. We'll see,
1:16:35 but that's my instinct. And so the the
1:16:37 other thing that I talked about was I
1:16:39 think it is a fine time to start
1:16:41 exploring uh open source. And I've never
1:16:44 been one to to push it because I think
1:16:46 it's complicated and overly technical.
1:16:48 But I think that the open-source tools
1:16:51 like LM Studio um or or O Lama where
1:16:56 where it makes it really easy for you to
1:16:58 install models and play with models just
1:17:00 like it's chat GPT. Those tools have
1:17:02 gotten pretty intuitive and pretty good
1:17:05 and then the hardware is getting good
1:17:07 enough to run these models that are that
1:17:09 are quite powerful. Um it seems to me
1:17:12 it's a good time to to start exploring
1:17:14 that. Um sorry for the all caps. had to
1:17:17 spell something out. Let's see. Had a
1:17:19 question earlier. I think it got lost.
1:17:22 Brandon, not sure how to reply. Do I
1:17:24 have to use the at symbol to type it?
1:17:27 I use voice to text so it gets weird.
1:17:30 Would love to learn how to mod here.
1:17:33 Cool. So, Cam definitely connect with
1:17:35 Brandon
1:17:37 at mention is another feature request.
1:17:40 The Streamyard chat is new. Okay, cool.
1:17:44 No, Emmy and Mary. It's been over a year
1:17:46 since I modded. It took me a while to
1:17:48 stop longtapping people's comments. Hey
1:17:51 Brandon, what's up?
1:17:52 >> Hey Kyle. So, so one thing I've I got to
1:17:55 say is I I'm getting whiplash from
1:17:57 looking down at my phone looking for Tik
1:17:58 Tok comments and Tik Tok pins [laughter]
1:18:01 because for the past nearly two years,
1:18:04 that's where a portion of our our
1:18:06 comments have been. I will say that this
1:18:10 is a work in progress, not only for us,
1:18:12 but also for StreamYard. the ability to
1:18:14 have this live interactive dialogue
1:18:16 inside of the inside of our experience
1:18:19 here is relatively new for them. It's
1:18:21 still basically a beta product that
1:18:23 they're not calling a beta product
1:18:24 anymore. So, we've got an active line to
1:18:27 them to say these are the feedback that
1:18:29 we're getting from our community.
1:18:31 >> Great.
1:18:31 >> This is what's making it difficult for
1:18:34 uh our users to engage with us. So
1:18:36 things like pinning, app mentioning. So
1:18:38 right now, if you see something, say
1:18:41 something to me, send me a DM in the
1:18:43 salon. I'd be happy to pass along and
1:18:45 elevate those messages up. Uh but we do
1:18:47 want to make it as easy for you to
1:18:49 engage with us as possible. So wherever
1:18:52 you're posting, uh we will find it.
1:18:55 >> That's great. That's great. The the
1:18:57 other thing that I wanna I want to point
1:18:59 out, um
1:19:04 let me see.
1:19:07 in the community feed.
1:19:12 Let me share my screen. Oh, I was
1:19:21 in the community feed.
1:19:24 I just did a prompt. Um
1:19:29 I just did a prompt. Um will you join me
1:19:31 in this discussion? So, for the past
1:19:33 three weeks, I've been out doing a lot
1:19:34 of different talks and um here on the AI
1:19:39 learning lab live, um this chart was the
1:19:42 direct result of a conversation. It was
1:19:44 a question that Kelly Camp posed when I
1:19:47 had said, "Things are about to get weird
1:19:49 a few too many times where she said
1:19:51 enough, Kyle. Like, what's it going to
1:19:54 look like? What does weird look like?"
1:19:56 And and this idea came out of it. I've
1:19:58 been presenting this out in the world
1:20:00 with a bunch of other ideas like the AI
1:20:02 readiness, the cycle of AI readiness.
1:20:04 I've been talking a lot about the salon.
1:20:05 I've been talking a lot about the great
1:20:07 repurpose.
1:20:09 This particular idea, the seven
1:20:10 economies has been really resonating.
1:20:12 Um, what it's allowing people to do is
1:20:15 kind of understand what's happening and
1:20:17 why it why it seems weird that some
1:20:19 people are so up on AI and other
1:20:21 people's are, you know, feel completely
1:20:23 disconnected from it. And it's allowing
1:20:25 them to to sort of check in and sort of
1:20:27 figure out where they are. Like you know
1:20:29 in the mall the the the red dot you are
1:20:32 here. Um and you know where do you want
1:20:34 to be? I want to go there. Um so I
1:20:37 started a conversation and I I I took uh
1:20:41 emails from a lot of people that I've
1:20:42 met over the past 3 weeks and I I sent
1:20:44 it out into the world and I said please
1:20:46 come into the into the community and
1:20:49 join this conversation. So, my request
1:20:52 to all of you now that I've I've got you
1:20:54 here and we're already in the community
1:20:56 is head on over in the lefth hand side.
1:20:58 Um, you know, the the channel that
1:21:00 you're in now is the AI learning lab uh
1:21:02 live irregulars channel and and four
1:21:05 spaces down is the community feed and
1:21:07 it's the sort of second the second uh
1:21:10 post in that channel. Um, do me a favor
1:21:12 and jump into that into that
1:21:14 conversation and just, you know, let me
1:21:16 know sort of where you are in the seven
1:21:18 economies. What I'd like to start doing
1:21:21 is is making the salon, you know, very
1:21:24 much what the original intent of salons
1:21:26 like in the 1800s was where we, you
1:21:29 know, sit down and have thoughtful
1:21:30 dialogues about about ideas. And so this
1:21:33 is just one I've been putting out in the
1:21:35 world that's been resonating. And so you
1:21:37 all are I I I this is not blowing
1:21:42 sunshine up your ass. I promise. I've
1:21:44 been out in the world a lot. the people
1:21:46 in the AI salon right now understand
1:21:50 what's happening with AI on on such a
1:21:53 deep level compared to where a lot of
1:21:56 the people in the world are that I think
1:21:59 you weighing in on things like this and
1:22:01 sharing your ideas on on just where you
1:22:03 are and what's going on um will actually
1:22:06 make a huge difference for people. So So
1:22:08 is this the homework for tonight? I
1:22:10 didn't uh think about assigning
1:22:12 homework, but yeah, this would be great.
1:22:14 If you want to take on a homework
1:22:15 assignment for tonight, jump over to the
1:22:17 community feed, find this prompt that
1:22:19 says, "Will you join me in a
1:22:20 discussion?" Uh, you can read my post
1:22:23 and I put a video there of me just uh
1:22:25 explaining this on on one of the lives
1:22:28 last week. Um, and then I also put a
1:22:30 link to the LinkedIn article if you want
1:22:31 to read more about it. and then just
1:22:33 jump in and just let let us know where
1:22:35 you are and where you want to be and
1:22:37 what your thoughts are on this and all
1:22:38 that sort of stuff would be great.
1:22:42 Okay, with that
1:22:45 um
1:22:47 for tomorrow's Oh yeah, that's a good
1:22:50 idea. Hang on a sec.
1:22:53 [clears throat]
1:23:00 Okay, back to the AI salon
1:23:05 um for tomorrow. So, we're down here in
1:23:09 the in the AI learning lab live. I
1:23:11 should see myself here, right?
1:23:14 Oh, click to watch live. I have to click
1:23:16 it. Okay, [snorts] look at look at AI
1:23:18 me. [laughter] I look so so bitching in
1:23:22 my fake leather jacket. Um, if you
1:23:24 scroll up, um, in this little top
1:23:27 section here and click on events, um,
1:23:30 then you'll see AI salon presents and
1:23:32 just click on that and RS RSVP for it.
1:23:35 All right? If you'd be so kind and then
1:23:37 come tomorrow, come join us. Um,
1:23:42 and with that,
1:23:44 I think I'm going to call tonight a
1:23:46 grand success. First of all, for all of
1:23:50 you who have been here from the
1:23:52 beginning and you know, sort of being
1:23:55 patient as we sort of, you know, work
1:23:57 our way through this thing that we're
1:24:00 doing. Thank you for your patience.
1:24:02 Thank you for coming to hang out here.
1:24:04 Thank you for modeling um what it means
1:24:06 to be curious and adaptable um for other
1:24:10 people. And for those of you that are
1:24:11 new here, welcome. Um I hope you're
1:24:14 enjoying your time in the AI salon. I
1:24:17 hope it gets better and better and
1:24:18 better and better. That's the intent of
1:24:20 this. That's why I put in all this time.
1:24:23 Uh and I really really really really
1:24:26 um deeply appreciate each and every one
1:24:28 of you. So for those of you who have
1:24:30 been mods and and can help with that in
1:24:33 any way, great. If you can't and and
1:24:36 can't figure it out, that's fine, too.
1:24:38 Um I hope you keep coming back and and
1:24:41 keep engaging. So with that, Kelly Camp
1:24:44 big success tonight. Yep. O Lama makes
1:24:46 it really easy. If you're new, we got
1:24:48 you. That sentiment by Lori Blair right
1:24:51 there.
1:24:53 I would say if there's a thing I I try
1:24:55 not to I try not to get prideful about
1:25:00 things, but one of the things that I'm
1:25:02 really proud of in the AI salon is that
1:25:05 it really is a welcoming, generous
1:25:09 community. And so if you come in here
1:25:12 without an absolute clue about what to
1:25:14 do with AI, this is a community that's
1:25:17 like, "We got you. We'll we'll point you
1:25:18 to the right right area." So if that's
1:25:21 you, if you're just like, "I'm going to
1:25:23 lurk. I don't want to say anything."
1:25:26 This is a community to say something.
1:25:28 Connect. Share what you're learning.
1:25:30 Kelly Anderson, what's up, new people?
1:25:32 Exactly. Um,
1:25:36 yeah, this is
1:25:39 it's a it's a really really special
1:25:42 community. So, if you're new here, um,
1:25:47 what you're probably not going to get
1:25:48 here is deep, in-depth tutorials and
1:25:52 prompt guides and things like that.
1:25:54 There may be some of that floating
1:25:55 around in here, but what you're really
1:25:56 going to get in here is people that
1:25:58 actually give a damn and give a damn
1:26:01 about you and making sure that you're
1:26:03 okay and you're finding your way. And
1:26:05 and here's the deal. We're all finding
1:26:08 our way. I have never felt more clueless
1:26:10 in my life as I do right now, this
1:26:13 instant. As as the as this stuff gets
1:26:16 more and more and more complicated and
1:26:18 more and more and more surreal,
1:26:21 I think it's important that we reconnect
1:26:23 with one another, that that we actively
1:26:25 go out of our way to keep connected.
1:26:29 Just in the last week, I've had several
1:26:31 aha moments that showed me how much I
1:26:33 actually know. Yeah, it was cool. Well,
1:26:35 that's the thing is we we all we we all,
1:26:39 as clueless as we feel, we've been doing
1:26:41 this for a while. Well, if you've been
1:26:42 in the salon for a while and you just
1:26:44 hang out here, part of the design of the
1:26:46 AI learning lab is not to actually learn
1:26:48 anything. It's to be in the
1:26:51 conversation. And so, by being in the
1:26:53 conversation, what you don't realize is
1:26:55 how much you're actually learning. And
1:26:57 so, what happens is you get in a
1:26:58 conversation with someone who's not been
1:27:00 thinking about this at all, and you
1:27:02 realize, oh, I actually know what's
1:27:04 coming. I actually know what's here. Uh,
1:27:07 so anyway, okay, Kelly Camp, I feel more
1:27:11 more lost than ever. To the clueless,
1:27:14 [laughter] to the three and a half year
1:27:16 clueless.
1:27:19 Clue the clueless unite. Um, no, we we
1:27:22 we do know more than we think we do. Um,
1:27:26 and you know, there there's a lot
1:27:28 changing really really quickly. Mary
1:27:30 Chester, I can attest to the wonderful
1:27:32 people in this in in this community.
1:27:34 Beautiful.
1:27:36 Beautiful.
1:27:38 Um, let's see. I've learned a lot here.
1:27:40 Beautiful. Thank you. That's really
1:27:43 awesome. So good. So good.
1:27:46 So so so so good. All right.
1:27:50 Happy Monday night. See you tomorrow
1:27:53 night at AI Salon Presents. Thank you
1:27:56 for everything and I will see you back
1:27:58 here in the salon Wednesday. So 710 out
1:28:02 in the world in the socials. So, if you
1:28:04 want to see the pre-show, that'll be out
1:28:06 there. I'll figure out what the pre-show
1:28:07 is. Like, I don't have a ton of thoughts
1:28:09 on the pre-show format, I'll figure it
1:28:12 out at some point, but but the whole
1:28:14 idea is the pre-show will happen from
1:28:16 7:10 to 7:30 Mountain, and then we'll be
1:28:18 in here 7:30 on Wednesday doing some
1:28:21 sort of demo, some sort of tool. I'll
1:28:22 figure out what it is. If you have any
1:28:24 requests for Wednesday, let me know and
1:28:29 maybe I'll uh I'll [laughter] we'll take
1:28:31 requests from the from the hotline. All
1:28:34 right, cool. Beautiful. AI is like golf.
1:28:37 My skills seem to come and go. Yeah,
1:28:39 exactly. Um days when you want to dive
1:28:42 head first and there's days when you
1:28:43 just want to watch. Yep. Exactly.
1:28:45 Absolutely. That's listen,
1:28:49 one of the things I'm learning in the AI
1:28:50 salon mastermind practice is that
1:28:55 slowing down and checking out
1:28:59 as an energy regenerator is actually
1:29:02 really powerful because you can come
1:29:04 back to the AI with objectivity
1:29:08 and actually see what's going on in in a
1:29:10 more clear way.
1:29:12 All right. Oh, and I don't have to limit
1:29:14 my shirts now. I don't I can go with my
1:29:16 cool shirts that are not going to
1:29:18 distract you cuz the comments are not
1:29:20 trying to compete with my patterns. So,
1:29:22 I can go full pattern, [laughter]
1:29:25 full pattern, full hair.
1:29:28 All right, everybody. This was awesome.
1:29:30 Thank you so much. Um, please keep
1:29:32 participating in the salon in between
1:29:34 these moments. Peace out. Uh, I hope you
1:29:37 had fun tonight. Bye.