The alpha, twice a week
That was 3 min.
Get the best new moments in your inbox every Wednesday and Saturday. It's free.
Check your inbox
We sent a link to your address. Click it to finish signing up. Not there in a minute? Look in spam.
or
Training Data · 3 min · Apr 2, 2026
AI works better as input than as a finished output
Jack Dorsey argues that people should treat AI recommendations as input to their own work, not as finished answers. He says judgment and a distinct point of view matter as copycat products become easier to make.
CEO, Block
Co-founded Twitter and Square, and leads Block, which is reorganizing around an AI intelligence layer.
Sonya Huang, Host · Partner, Sequoia Capital
Works with OpenAI and Anthropic; co-wrote Sequoia's essay on generative AI.
Pat Grady, Host · Co-steward, Sequoia Capital
Joined Sequoia in 2007; backed Snowflake, Zoom, ServiceNow and OpenAI.
About this segment
Jack Dorsey argues that people should treat AI recommendations as input to their own work, not as finished answers. He says judgment and a distinct point of view matter as copycat products become easier to make.
From “Jack Dorsey: Every Company Can Now Be a Mini-AGI”, Apr 2, 2026. Watch the full episode
Thanks. We'll look at and fix what's off.
“most people are seeing what the intelligences do, these tools do as output rather than better input to create better output ourselves.”
Jack Dorseyat 1:00:04 · Training Data1:00:04Takeaways
- Treat model recommendations as inputs, not finished work.
- Question your assumptions instead of defaulting to the tools’ suggestions.
- Use taste and a distinct point of view to avoid copycat products.
In this episode
Full episode- 30:58Block set its workforce size by rebuilding the company on paper5:52
- 7:02Dorsey wants Block to operate with three distinct roles6:22
- 48:17A founder may need to lose credibility to build a second act2:52
- 13:45Transaction data lets Block anticipate customer needs3:28
- 37:10Founders should choose investors like key hires4:13
- 42:16Overdelegating left Block operating like a holding company2:44
- 44:59Cheap prototypes make broad exploration more practical3:21
58:31AI works better as input than as a finished output3:09
Transcript
Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.
- 58:33
on that like when you're you've hired some amazing people in your career that have gone on to do some amazing stuff what do you what do you think and what's changed what's what's timeless what's changed i value someone who's able to reprogram their mind and assumptions constantly um i mean this is stated a lot but just being able to question your own requirements your own assumptions and your own decisions or how rigid you are uh towards your past the company's past and sacred cows or what the competition is doing um it's just you know being able to like go wild for a bit and then also being able to get that entire corpus down to something that's actually manageable and that can be articulated in a way that other people understand.
- 59:22
I think that's extremely valuable. Has it been valuable or is it more valuable because things are moving so fast right now? I think it's more valuable. Yeah, I mean, I think it's going to be so easy to go along with the momentum of what's happening around us today. And it's going to be increasingly hard to break free of that momentum given what the tools do and this is the way we do things. And I think people will, because we're offloading some of our intelligence to intelligences and people will go along with more likely to default to.
- 1:00:00
what these things are suggesting rather than seeing them as an input. Like I think we're still in the mode where most people are seeing what the intelligences do, these tools do as output rather than better input to create better output ourselves. And like I think that's important to me. Like when I was doing that like every day, three hours, like this is input to me. And it's now up to me to like really make better output from like all this new input that I have and like the ability to just take all this and see it at once it's just phenomenal so having someone who's able to discern signal from noise and cool from not relevant you know like there is there is this is highly overused right now but um the taste thing is real and it's it's not just like i know what things look good together it's more like do i have a point of view do i have a
- 1:00:58
perspective and an opinionated drive to get it there and is it is it relevant is it more relevant than than you know what else is out there and i think that's that's critical that's what any founder is doing is like i'm building something that didn't exist in the world because i wanted to see it there and and because it didn't exist is why i'm building it and i think um right now you're seeing a lot of companies they're just copies of copies of copies of copies because it's easy you know instead of like what's your what's your point of view like what is what is opinionated in this and like what where is it pushing the the boundaries and where is it uncomfortable i think that's brilliant it's great place and thank you both for coming on long