AI is full of hype. Here’s who I actually follow

One of the principles I follow when

learning AI is to follow builders not

influencers. There's so much hype around

AI. Like every day on X, all these

influencers are claiming it's over or

this tool is insane. A lot of

sensational claims trying to grab your

attention. But I found that the best

insights often come not from influencers

but from builders. People who are

actually making AI products as their day

job and who post about their process of

building AI products on the side. I

define these people as builders. So

these can be people who work at the

model companies like Open Eye Anthropic

or they could be startup founders

working on AI applications or product

managers of AI products that I really

admire. I define influencers as people

whose full-time job is to post content

online and make money from doing that.

Here are some of the specific builders

that I've come to admire and I follow

all of their work and their sharings

online. The reason I prefer to follow

builders is one, they actually have

hands-on experience building with AI.

they give a lot of actionable advice

from their real world experiences.

Number two is they're usually more

authentic because they don't make a

living through making content whereas

for influencers a lot of them make money

through sponsorship. So they will

promote a lot of AI products and tools.

But you never know if they're promoting

it because they genuinely like the tool

or because they're just paid to say good

things about them. Sometimes I would

watch a video by AI influencers and

think to myself, do they really use the

product or do they just dabble in it

because some AI starter pay them to talk

about it? The specific thing I do to

follow builders is to use YouTube as a

search engine for their insights. So for

example, one of the product leaders I

really admire is Chris Padrigo, who is

the founder of Granola. I would put his

name in the YouTube search bar and I

would basically watch all of his video

podcasts or interviews or speeches that

he's done in the past. For example, like

this one where he talked about how he

decided on what to build and what not to

build.

>> Text uh window length. It was too small

when we launched. You could only do like

30-minute meetings.

>> Um and we could have done a bunch of

work to like okay try to chunk that or

we just just wait a little bit and the

context windows got keeps getting

bigger.

>> Yeah, exactly. Yeah, this tactical

advice is super helpful as I think about

building my AI product and it could only

have come from someone who had hands-on

experience with building. One of my

favorite things about Silicon Valley is

this open sharing culture where people

love talking about how they're building

what they're building on platforms like

X and YouTube. So one process that I

found to have worked for me is first I

would identify AI products that I really

enjoy using and then go and find who

built this product like who is the PM or

the founder behind it and then I would

go on X and YouTube follow their Twitter

watch all of their video podcast

interviews and conference speeches and

what they say makes so much more sense

after I actually use their products

because when you hear about how they

share the behind-the-scenes thinking

behind all these little product

decisions, it starts to click for you.

And this for me has become the most

effective way to learn how to build an

AI.