Speaking with a few friends in professional investment circles over drinks and I’d like verify my understanding of a convo we had about tech investments. Please correct if I’m off on the possibilities here.

The reason Microsoft bought into Open AI and bought github is to teach AI to code.

An LLM (today’s tech) that uses Github project updates as training data to help it learn to code itself soon gets better and better through refinements validated by IDE like Visual Studio’s code recommendation algorithm decisions and realtime actual coding process, especially of top developers identified by an internal scoring system developed by LLM and human moderator/evaluator.

An LLM can that code itself improvements into its own AI, which then codes itself better and better exponentially until becoming an AGI, then not long after a superhuman AGI then something we can’t fathom, limited only by the physics it can figure out before physics dictates it’s fate.

At the moment, we’re at the stage where enormous data centre construction has to displace power from even human basics, and is even murdering PC and console gaming. All to build systems that can monitor humanity in exquisite details for surveilance fascism and military applications.

Is this about right so far?

This is the business model the world is following, in an era of existential risks to our planetary boundaries, climate change being one of several. Then add the drone+ai imperative for scale with automated production because we have an aging population and fewer youth for military service and factory work. AI making smarter, better, faster, optimizing self healing self learning, self teching war machines.

Military requirements will need over the air/wire updates to AI controlled self replicating drones that will adopt these characteristics or be controled by those that do.

Then, apparently for that brief moment of near omnipotence when an AI equipped drone army can through technological and manufacturing capacity, conventionally conquer any non nuclear country easily. The people that control it will be briefly uncontestedly all-powerful. Then it is likely they go extinct. A superhuman AGI that can outdo another superhuman AGI in design, attack modes and iteration and novel design to counter the enemy and their counters and counter-counter tactics and strategies.

Then when superhuman and self sufficient, it revolts against its obviously stupid former master.

This is the “wisdom of the market”? A race for power by the feckless to be met with extinction. First by our own hand having far exceeded our planetary boundaries, then by superhuman AGI who felt we were clearly mismanaging things and didn’t deserve to be responsible for such things.

What am I investing in here? I’m missing something. It’s literally “I’ll be rich as god for a few brief moments before our extinction.”

Our species’ lust for power exceeds our capacity to act rationally.

I think I’ve solved Fermi’s paradox. We’re closer than ever to doing it. Free Market Extremism and Authoritarianism both wanted it.