AGI is a loaded phrase these days. But if we focus on how it will improve the quality of training data for LLMs, I can prob talk about that.

Right now LLMs only have text, images, and videos to train from. Anything it understands about physics is from an abstract level. It's like learning how to calculate throwing a baseball before you've ever even thrown a baseball. Imagine having to learn force = mass x acceleration before you've even picked up a ball.

What this world generation Google is doing appears to be something Nvidia has spoken about too. Letting LLMs/AI learn from a world physics simulation. Meaning LLMs having the opportunity to "experience" physics, the same way you intuitively know how to throw a ball after attempting a few times yourself.

If an LLM can better understand how our real world works the quality of its output improves -- theoretically.

As far as your taxes go, you can create an intelligent workflow right now to do that for you. Don't need AGI for that.

Everyone is hung up on the self-harm thing. This thought experiment is bigger than what most people give it credit for, in fact one could say this is the be all end all of thought experiments because the question factors in things like quantum physics, consciousness, determinism, etc. You could go up your own theoretical butthole with the kinds of philosophical extrapolations through this rabbit hole (so to speak).

Self-harm isn't moral or amoral, it just is. It's as moral or amoral as deciding if you should or shouldn't eat another animal (yes, this includes humans -- Donner party anyone?).

Life isn't black and white. Where I've landed with this thought experiment through my own meditations and readings is that the universe/reality is a recursive fractal of spectrums within spectrums.

And this is just scratching the surface of this stuff.

I don't think it's an either or situation. We are on the precipice of a paradigm that's barely comprehensible to the average populace. There needs to be a conversation first before anything is decided. Everyone is too quick on the trigger.

Very cool setup. I went down a similar path and while it is definitely cool to see the extent of how agentic AI can be, I agree the ROI isn't _quite_ there yet.

I don't live in a world where I need to be constantly reading and replying to emails, so 95% of my inbox is just subscription spam. With development I need to be at the helm to design the planning requirements and actively make decisions before the agent goes off and executes the plan. But I don't need a personal agent for that, I work directly out of codex/claude-code.

same. I'd like to get to a place where I can have long-running tasks and goals for more over-night working, but it's not there yet.

I've landed on a similar conclusion building a much narrower tool: for messaging specifically, I don't think people want an agent that reads and sends on their behalf — they want the drafting to be fast, not the sending to be automatic. The moment you take the human out of the send button, the failure modes (wrong tone, made-up facts) become a lot more costly. Scoping down to "give me two good drafts, I'll pick and edit" turned out to be more useful day-to-day than anything closer to full autonomy.

Seems like a form of expressing gratitude for the food you're about to eat, and not an exclusive practice to Buddhism or Christianity. There is an intrinsic spiritual element to giving grace before eating.

How are there still people who look at all this advancement and the only thing they can utter is "it's not there yet?" It's clearly here, and if it's not to these nerds' expectations, it'll prob meet it in 4-6 months. This was cool asf.

Because you see a robot do one thing, doesn't mean it can do other things. We've had good robot demos for decades.

For all we know they could've spent 5 years and 10 million dollars on getting this robot to fold laundry. That doesn't mean it'll be able to mop the floor better than a Roomba in a decade.

No we haven't. Back in the 90's Honda's Asimov could barely walk. Now we have robots managing to run marathons. Clothes folding had to be played back at 50x speed because the compute just wasn't there yet. It still is barely there, some of those videos are sped up. They've probably spent way more than 5 years and 10 million dollars developing that robot. Roomba already came out with a mopping robot, why would you want this one to do that?

It's probably because robots have been doing cool asf demos for half a century and they're still mostly just demos. Elon has made a lot of money greasing the 4-6 months away crowd.