quonn

Elon could turn Tesla's car business around by apologizing, letting go of his extreme compensation claims, keeping shares of only 15% instead of 25% and working pro bono in the future.

Otherwise it is difficult to see how they could still succeed, even if they ever manage to produce a perfect FSD.

He brought this on himself.

Ukraine was not in the examples given and you read both the original post and my response wrong. The post did not say „if you have nuclear weapons you don‘t get attacked“. It said „if you don‘t have nuclear weapons you get attacked“. Of course you get attacked by Russia either way, but that‘s not the point.

The point is that you don‘t get attacked by anyone else unless you‘re a dictatorship.

How do you get the idea that we have any of that technology? Brains are nothing like computers and we have almost no technology that is like biology. There is no reason to be sure that we can just scale down the existing tech to arrive at something as efficient. Maybe we can. But if not then developing a system that‘s more like chemistry could take 50 years or 100 or more.

A bunch of matrix multiplications is not a species.

> You lack that deterrent you get attacked, as simple as that

No you don't, unless you're a dictatorship (including all the examples you gave).

What's wrong with TypeScript for backends? Specifically backends where the load is moderate or happening in the database anyway?

> When you have exponential growth, it's always early days.

Let‘s take the development of CPUs where for 30-40 years the observable performance actually did grow exponentially (unlike the current AI boom where it does not).

Was it always early days? Was it early days for computers in 2005?

I think you severely underestimate what the 8 billion human beings on this planet can and will do. They are not like horses at all. They will not allow themselves to be ruled by, like, 10 billionaires operating an AI and furthermore if all work vanishes then we will find other things to do. Just ask a beach bum or a monk or children in school or athletes or students or an artist or the filthy rich. There _are_ ways to spend your time.

You can‘t just judge humans in terms of economic value given the fact that the economy is something that those humans made for themselves. It‘s not like there can be an „economy“ without humankind.

The only problem is the current state where perhaps _some_ work disappears, creating serious problems for those holding those jobs.

As for being creative, we had GPT2 more than 5 years ago and it did produce stories.

And the current AI is nothing like a human being in terms of the quality of the output. Not even close. It‘s laughable and to me it seems like ChatGPT specifically is getting worse and worse and they put more and more lipstick on the pig by making it appear more submissive and producing more emojis.

> It's early days and nobody knows how things will go, but to me it looks that in the next century or so

How is it early days? AI has been talked about since at least the 50s, neural networks have been a thing since the 80s.

If you are worried about how technology will be in a century, why stop right here? Why not take the state of computers in the 60s and stop there?

Chances are, if the current wave does not achieve strong AI the there will be another AI winter and what people will research in 30 or 40 or 100 years is not something that our current choices can affect.

Therefore the interesting question is what happens short-term not what happens long-term.

Given how Japan works in general I bet it's the latter. It's a great country to travel and eat alone, for example.

> merge the same concept

It's doing compression which does not mean it's coherent.

> The addition circuits are also fairly easy to interpret.

The addition circuits make no sense whatsoever. It's doing great at guessing that's all.

> that they can internally represent many different complex ideas efficiently and coherently

The Transformer circuits[0] suggest that this representation is not coherent at all.

[0] https://transformer-circuits.pub

In the end the data has to fit into structures or tables that can be processed by some algorithms. If the system is not rigid to a certain degree it would become unmaintainable or full of bugs or both.

> The person lamented the ongoing economic volatility — caused by tariffs and Trump’s unpredictability — during a presidency that they had been promised would be a boon to business. “We’re all experiencing a liquidity crunch,” they said. “We need public markets to open.”

"Promised", boohoo. These people really don't understand what a society is and are operating at the level of toddlers to put it mildly.

If they ever manage to destroy many jobs by building strong AI, we will hear "We're experiencing a lack of demand".

> most CS has now become blue-collar work, and it is something people in the industry have not yet come to terms with.

No, actually it's literally office work, working in an office or even working from home.

How is it designed so? You pass a law that puts anyone running a node 10 years in prison and anyone accepting or paying through it as well. You see, easy.

For this to not be possible it would have to be widely in use and government-backed.

Why would it have anything to do with AI? Generative AI has been widely used for two years and the drop is exactly around January 20. What happened in AI around that time?

As if western europe wouldn't have the skills. Eastern Europe has not relevant international software industry outside of JetBrains. Same as western europe which at least has SAP and Spotify.

No, it wouldn‘t.

It may sound plausible, but the actual computations are very simple, dense and highly optimised already. The model itself has room for improvements, but this is not necessarily something that an engineer can do, it requires research.

If there were evidence that one LLM would be conscious I would also accept it for others. However this is not the case.

But we know at least one human is conscious. That‘s convincing me.

Excel and Powerpoint are not conscious and so there is not reason to expect any other computation inside a digital computer to be different.

You may say something similar for matter and human minds, but we have a very limited and incomplete understanding of the brain and possibly even of the universe. Furthermore we do have a subjective experience of consciousness.

On the other hand we have a complete understanding of how LLM inference ultimately maps to matrix multiplications which map to discrete instructions and how those execute on hardware.

Consider this: Being born/trained in 1900 if that were possible and given a year to adapt to the world of 2025, how well would an LLM do on any test? Compare that to how a 15 years old human in the same situation would do.

Given the fact that LLMs like most neural networks work by passing their input through layers, wouldn't this be expected? There's no going back to an earlier layer and if the first layers are in some sense needed for "translating" [0] to English, any other functionality in those layers cannot be used.

[0] I am simplifying here, but it would make sense for an LLM to learn this, even though the intermediate representation is not exactly English, given the fact that much of the internet in English and the empirical fact that they are good at translating.

For this the Russian buyer would have to previously sell something to the Philippines and accept pesos. Why would they accept those pesos if they are not generally accepted elsewhere?

True - but LLMs can do this.

A German Onion-like magazine has a wrapper around ChatGPT that behaves like that called „DeppGPT“ (IdiotGPT), likely implemented with a decent prompt.

I can do this because it is in fact the most likely thing to continue with, word by word.

But the most likely thing to continue a paper with is not to say at the end „I don‘t know“. It is actually providing sources which it proceeds to do wrongly.

It‘s broken now. It was fine 5 years ago.

Not necessarily. There is no big difference whether the business logic resides in the same node process or another one. If the first process is unsafe on that level, then the token can also be extracted.

Maybe it’s like measuring the detonating fuse and give them matches:

https://en.m.wikipedia.org/wiki/The_Fire_Raisers_(play)