This is a really good feedback! It's possible that we need to massage some of the vocabulary. I've gotten this feedback from a few people now.
When we wrote this, an AI-first company in our head was a "stronger" version of an AI-native company. Sort of like how you can be a native speaker of a language, but not necessarily an expert of the language. The "-first" implies a sort of AI-oriented maximalism.
How are you thinking about AI-first in relation to your own company? What problems/opportunities do you see?
Ok that's good to understand! I will need to reread the whole piece now.
I assumed AI-native was the stronger version. Like Bain telling Batman "You merely adopted the dark, I was born in it."
I am quite excited about the shift to stochastic thinking. I invest in public markets, everything we do needs to be probabilistic. Excited to see where this new world leads
Love the range from technical to theoretical in this piece and how being an AI-first co asks humans to be empathic and flexible (the best of what’s distinctly human). I’m curious about your last line - “AI systems will eventually be as fundamental to our cognition as the psychological technology of literacy.” I’m sensing you mean that AI will be as transformative to humans as language has been… (one of my favorite insights from Vervaeke…) Would love for you to unpack this as you see it…
Thanks for the feedback! Yup that's right. We were trying to evoke the term "psychotechnology" from John without taking a discursive detour in introducing it. But yes, we believe that AI will lead to the emergence of novel psychotechnologies. And that these technologies will be at least as profound as the psychotechnology of literacy.
We don't 100% know what these look like. But I've started to get intimations of them. For example, various AI tools that change how groups of people organize themselves together.
Excellent write up, thanks Varun! I'm a huge believer in the 'business as a spec' (i.e. evals) idea you laid out here.
Really enjoyed this!
Re: this part: "An AI-native company is one that’s largely a deterministic-native company which has its workflows accelerated with AI-based tools."
Did you mean to say 'that's largely a stochastic-native company'
This is a really good feedback! It's possible that we need to massage some of the vocabulary. I've gotten this feedback from a few people now.
When we wrote this, an AI-first company in our head was a "stronger" version of an AI-native company. Sort of like how you can be a native speaker of a language, but not necessarily an expert of the language. The "-first" implies a sort of AI-oriented maximalism.
How are you thinking about AI-first in relation to your own company? What problems/opportunities do you see?
Ok that's good to understand! I will need to reread the whole piece now.
I assumed AI-native was the stronger version. Like Bain telling Batman "You merely adopted the dark, I was born in it."
I am quite excited about the shift to stochastic thinking. I invest in public markets, everything we do needs to be probabilistic. Excited to see where this new world leads
Ohh that's so interesting! Do you already use LLMs a lot in your workflows? I'm also curious how you found our substack?
Yes LLMs are already at a level where they can replace much of what a junior analyst could do. Already using a lot, want to use them much more
I followed Dan back in 2020 so the cross post hit my inbox
Love the range from technical to theoretical in this piece and how being an AI-first co asks humans to be empathic and flexible (the best of what’s distinctly human). I’m curious about your last line - “AI systems will eventually be as fundamental to our cognition as the psychological technology of literacy.” I’m sensing you mean that AI will be as transformative to humans as language has been… (one of my favorite insights from Vervaeke…) Would love for you to unpack this as you see it…
Thanks for the feedback! Yup that's right. We were trying to evoke the term "psychotechnology" from John without taking a discursive detour in introducing it. But yes, we believe that AI will lead to the emergence of novel psychotechnologies. And that these technologies will be at least as profound as the psychotechnology of literacy.
We don't 100% know what these look like. But I've started to get intimations of them. For example, various AI tools that change how groups of people organize themselves together.
And yes, future essays will explore this!