00:00:00
The moment I had where I became clawpilled, I was like, "Hey, I want to build my network. I know these 20 people
00:00:06
in Japan. I had dinner with them during my recent trip. I want to know who they know. So, check out LinkedIn and other
00:00:11
things and who they're associated with and make me like a mind map of it. And
00:00:14
then the next trip, I want to meet with the next circle of those connections."
00:00:19
She's like, "Okay, I got the results." And they said, "Where do you want me to
00:00:22
put them?" I was like, "Well, where can you put them?" And it said, "I could put
00:00:24
it in a Google sheet. I could put it in a notion table. I could put it here. I could give you a PDF. I can give you a
00:00:27
CSV file. or I could write you a CRM. And I was like, "Yeah, sure. Make me a
00:00:31
CRM system." And it made me a CRM system. I think maybe one out of a thousand people working with AI have had
00:00:37
that experience. Maybe it's one in 10,000 where your agent says, "I'll make
00:00:42
you bespoke software. The context is so good because the memory is getting better." Yeah.
00:00:47
>> When did that happen? >> I think it it happened with Opus4. That was the inflection point when models
00:00:54
started being amazingly good at orchestration and reasoning and tool calls. the model started becoming very
00:00:59
good at handling the context. So the context window no longer became a problem and that made it like suddenly
00:01:05
so good at doing very long orchestration tasks.