37) What Do I Think?

Behind-the-scenes building Vambrace AI, a company on a mission to figure out its mission. Please pardon the stream-of-consciousness style. Subscribe to follow along or visit the site here:

(typos are to make sure you’re paying attention)

Introductory Remarks

Dear Vambracers —

In last week’s post, Welcoming Walter, I reflected on my early experiences with clawdbot (now OpenClaw) and explored near-term use cases for the always-on, fully-powered agent. In the past week, Walter has helped me: build out a massive lead list for a new incubation, kept track of ongoing tasks as they occur, reviewed call notes, calendars, and emails, to keep the list of tasks current, joined moltbook (Reddit for bots), and built me a complete GTM playbook for the previously mentioned incubation. And all of this was accomplished from the comfort of my phone. The experience is not without glitches and hiccups, but such is life with an emerging technology. And I’m just starting to glimpse the potential that exists at the confluence of AI, data, and autonomy.

Data is the moat

I’ve been incredibly excited about AI’s potential within business software, given its near-term ability to basically review, advise on, and or run many aspects of a company (pretty much anything involving data). But my extreme excitement can cloud the logic that underpins my beliefs—and so today I want to articulate a more calculated and reasoned basis for my excitement.

Foundational beliefs

AI & Data

• Artificial intelligence (AI) is fundamentally a data transformation technology. Given some set of inputs, AI leverages predictive models to generate a likely set of outputs.

• Virtually everything can be or is expressed as “data.”

• If you assume any rate of improvement for AI, then it will eventually be one of the most transformative technologies of human history.

• If you assume any rate of continued decline in cost of compute, then AI will eventually be one of the most abundant technologies of human history—limited mainly by our energy generation capacity.

• AI output is only as good as the data input. Proprietary business data allows for unique AI-generated business insights.

• Since AI is “machine-generated” it is uniquely good at understanding the code and language that animates machines.

Data & Software

• Most modern businesses run on data. The software that powers any platform is just massive sheets of interrelated data, presented in a clear way.

• Most “modern” software was made for humans and is therefore optimized for human-computer interaction, which most often results in point-and-click.

• Being made for humans, most modern software abstracts away the complexity associated with the code and the database structure of a given software.

• Data is effectively context. Any point-in-time capture of a company’s database represents that company’s reality at that given moment.

• High fidelity and readily accessible data will be critically important for companies to leverage AI tools to drive business intelligence and eventually take action on behalf of the company.

Software & Humans

• Within business, humans are fundamentally data transformation agents. Given some set of inputs, we apply lived experience, judgement, constitution, etc., to drive some set of outputs.

• There are many data transformation tasks within modern business that don’t require massive amounts of experience, judgement, constitution, etc.

• Software is generally emotion-less; humans are emotion-full.

• The current point, click, and type interface of human-computer interaction is probably not the optimal long-term interface.

Humans & AI

• Humans thrive on human-to-human interaction; empathy is what makes us human.

• As AI improves and accelerates and diffuses, the only zero-trust way to verify something is real will be to experience it physically (at least in the near-term and assuming we are in shared base physical reality).

• If we can eventually “AI away” the day-to-day functioning of society and industry, then that will free humans up to pursue other things. It’s currently unclear to me (and most people) what those other things might eventually be.

• Humans are fearful of and resistant to dramatic changes to the status quo (s/o High School Musical), but change is inevitable.

• Humans are resilient and can exhibit extreme ingenuity when faced with a shifting world and technological order.

• Nobody really can say with any certainty what the world will look like in 5, 10, 50 years. I intend to judge expert commentary based on the perceived integrity of the expert.

• Economics are inextricably linked to human behavior and technological development.

So, now what?

I’m not sure exactly where to go with this. I think the purpose of today’s post really was just to put finger to key on some things that I believe to be true. There’s so much noise in the popular commentary surrounding AI, and things are changing so quickly, that it can be hard to give myself the space to think and reflect.

Looking Forward

I believe that any productivity-enhancing technology will eventually be a net-positive for human civilization, but I acknowledge that there will definitely be some near-term hardship. Growth is hard, and it’s impossible to experience any real change without some pain.

But humans are resilient and I’d bet on us 100 times out of 100. And I don’t want to live in a timeline where I’m not adopting an optimistic outlook, rooted in my belief in the innate goodness of people, even if it seems naive.

Have a wonderful week!

Sincerely,

Luke