AI Learning Lab LIVE Clip Grid! 9 Clips from 8.29/25

Description

Chapters

Transcript

0:01 I don't know how to use it.
0:04 Like I'm getting really inconsistent
0:07 results. Sometimes I get results that
0:09 are like mindblowingly good. Sometimes I
0:12 get results that are like it didn't even
0:14 try to give me what I thought it would
0:16 give me.
0:18 Um I feel similarly with chat GPT5 right
0:22 now. Still I feel like a clueless
0:23 newbie. Feel like an idiot. And I don't
0:27 listen I idiot might be a harsh word.
0:31 absolutely reminds me of when I was
0:33 following NFT projects. I was like, it's
0:36 like, oh my god. And and they're just
0:40 relentless. And I and I feel like it's
0:42 the same way with these things. Like
0:44 they just keep, you know, releasing new
0:46 things. So
0:49 he's only following 493 people. So you
0:52 can click on that and you can now go see
0:54 the people that he follows.
0:57 And
0:59 sometimes you'll find people in here
1:01 that you're like, "Oh, I should follow
1:02 them."
1:04 Saw a LinkedIn post yesterday that said
1:08 10,000 prompts is the new 10,000 hours.
1:12 And basically meaning like if you want
1:14 to get good at a tool, you just got to
1:16 [ __ ] use it. you
1:20 tech intakes language and outputs the
1:22 average of that language nothing more.
1:24 That's not true. That's not true.
1:27 That's not true. That is a trope from if
1:32 you look at how
1:35 a large language model
1:38 predicts the next possible token.
1:43 And we mash together those disperate
1:45 thoughts in this crazy straw of a
1:48 [ __ ] brain we have and we go, "Huh,
1:51 here's a new idea."
1:54 Which is some weirdass combination of
1:57 the thoughts that got put in our head
1:59 over our lifetime. How is that [ __ ]
2:01 different than what a large language
2:03 model does?
2:06 So imagine being so fil with AI tools
2:11 that you could just sit there like the
2:14 the Wizard of Oz or like the the
2:17 conductor of a symphony
2:20 and just go, "Okay, I'm going to pull in
2:21 this tool and do the data analysis. Then
2:23 I'm going to take the answer of that.
2:24 I'm going to put that into a
2:25 presentation and this afternoon I'm
2:27 going to give that presentation to the
2:29 boss."
2:31 Like that study is [ __ ] ridiculous.
2:34 And so CEOs that expect a 40% bump in
2:38 revenue because we gave people access to
2:41 J chat GPT is [ __ ] [ __ ]
2:44 Nobody knows how to use it.
2:48 What did it do?
3:00 It just wrote us an application.
3:06 It just coded us a website.
3:14 Holy [ __ ]
3:17 Okay. Do Do you remember when I said
3:20 before, I don't know how to use this
3:22 thing?