TheRealDarthMaul

[TheRealDarthMaul @TheRealDarthMau]

@elonmusk @yunta_tsai Good thing about X is it puts a timestamp on your ideas so that you can show-off "I said so before it was mainstream".

https://t.co/W7baivxtEW

[TheRealDarthMaul @TheRealDarthMau]

@alz_zyd_ Assume a perfect LLM (or another AI application) trained with the whole human knowledge up to 1905. Would it be able to generate Einsteins thought experiment of relativity by itself? If your answer is no then the answer is yes.

[The Scientific Lens @LensScientific]

It is by logic that we prove, but by intuition that we discover.

— Henri Poincaré

[TheRealDarthMaul @TheRealDarthMau]

@alz_zyd_ Exactly. The best part is you decide the format that you can consume easier. Table, bullet points, distilled info, summary, detail, correlation, etc.

[TheRealDarthMaul @TheRealDarthMau]

@alz_zyd_ Maybe a “simulation in a box” should be the eventual target of humanity where all bodily needs are fulfilled automatically, including reproduction. Would allow roaming the space for billions of years until the heat death of the universe. Work should continue until then.

[TheRealDarthMaul @TheRealDarthMau]

@kimmonismus What will be more interesting is when the models are injected in vivo.

[TheRealDarthMaul @TheRealDarthMau]

@kimmonismus Let me fix that for you:
Long-horizon tasks → autonomous agent societies → fully automated “no-person companies” → AI training AI → self-evolution → self-awareness simulation → (better) emotion simulation → (better) consciousness simulation → ASI (in slico).

[TheRealDarthMaul @TheRealDarthMau]

4/4 (Bonus)
Training discovers the important directions (think "unit vectors") of the N-dimensional conceptual space, while NN weights learn how directions relate. Inference is one forward pass through that geometry, not an exhaustive search. That's why LLMs feel remarkably human

[TheRealDarthMaul @TheRealDarthMau]

1/3
Even if you trained an LLM on everything humanity knew before 1915, it still wouldn't discover Einstein's relativity. Not because it lacks intelligence, but because the required concept isn't represented in the space it has learned.

[TheRealDarthMaul @TheRealDarthMau]

3/3
Before 1915, none of those conceptual directions represented the possibility that space, time, and simultaneity are not absolute. Relativity wasn't just another point in the existing space, it required a fundamentally new conceptual direction: an N+1 dimension.

[TheRealDarthMaul @TheRealDarthMau]

@thegautamkamath @CarinaLHong @MarkSellke @SebastienBubeck @WittedNote Hello, will this be on YouTube?

[TheRealDarthMaul @TheRealDarthMau]

The year is 250K.

Humanity is so strong that they survived a meteor impact by annihilating it.
But still consciousness, life, the very reason of existence are as mysterious as today. More than 250K years of written history but no significant progress.
How depressing would it be!

[TheRealDarthMaul @TheRealDarthMau]

@amandaorson Art, sports.

AI will attack to art. Possibly will be succesfull on most fronts. But live performed art by humans will survive. Like theatre.

Similarly human sports which fuel the tribal instincts. People will never watch robot matches as much as football.

[TheRealDarthMaul @TheRealDarthMau]

@ChuksUmeche @amandaorson @pmarca Hahahaha
This was good.

[Amanda Orson @amandaorson]

The faster technology moves, the more I think about Bezos' question

What won't change in the next 10 years?

Things I've been writing down over time:

- Humans will always need shelter, food, energy, and healthcare.
- The desire for ownership and the accumulation of wealth.
- The physical world will move more slowly than the digital one.
- Every increase in technological capability, especially AI, will require more energy.
- People and businesses will continue to need access to capital.
- Capital will continue to seek returns that exceed inflation.
- Underwriting methods evolve, but demand for credit (loans) is persistent.
- Trust remains scarce and becomes increasingly valuable as content, code, and fraud become cheaper.
- Verified identities and reputation becomes more important as information becomes abundant and synthetic.
- Long-term wealth creation and dynastic (multi-generational) thinking predate modern technology, and will persist.
- Coordination and transaction costs never fully disappear; market friction will continue to justify the existence of firms and intermediaries.
- People will continue to compete for status.
- Consumers will pay a premium for products and services that confer status.
- Time remains fixed at 24 hours per day.
- But attention is a finite resource and an enduring constraint.
- Products that credibly save time (or enable delegation) have a perpetual market.
- Inaccessible, proprietary data will be a persistent moat. The more inaccessible and difficult to aggregate, the deeper the moat.
- People want accountability, recourse, and clearly identifiable responsibility when things go wrong.
- Regulation consistently lags technological innovation.
- Compliance requirements, licensing, and regulatory moats persist even when machines can perform the underlying task.
- Local knowledge remains valuable and difficult to replicate.
- Heterogeneous markets (like real estate) continue to reward people with deep contextual understanding.
- Incumbent organizations tend to underinvest in disrupting their own businesses, which always creates opportunities for challengers.

Bezos' insight on what wouldn't change in 10 years was "Customers will always want lower prices and faster delivery."

It's boring/ true, but I think that's the point.

Everything we build today can and will be rebuilt more cheaply, faster by someone else.

Build on the invariants, not the trends.

What have I missed?

[TheRealDarthMaul @TheRealDarthMau]

@elonmusk @ThomasAlxDmy And you should not have a competition.

[Joel David Hamkins @JDHamkins]

Introducing the Lectures on Infinity, containing all my favorite paradoxes and conundrums.
https://t.co/dA9qDt9LG8 #InfinitelyMore #BookOfInfinity

[TheRealDarthMaul @TheRealDarthMau]

@rrwilliams You are doomed if it initiates the sky net.

[TheRealDarthMaul @TheRealDarthMau]

@yuhasbeentaken It is remarkable that GLM dethrones DeepSeek. I think it increase the credibility of the Chinese open-source AI ecosystem, being a relatively old spin-off from the @Tsinghua_Uni as one would expect, not a questionable “side project”.

[TheRealDarthMaul @TheRealDarthMau]

@rakyll Neither do I. Once upon a time I was sort of an hooligan.

And somewhere around 15+ years ago I lost my interest with football.

I think “rareness” of matches vanished and that was the reason I started disliking football.

Now I find women’s volleyball much more exciting.