red75prime

You just need to remember that garbage (unreachable reference cycles) isn't collected.

Or a single disease the New World inhabitants have adapted to. Someone could have been bitten by a bat or something like that.

Turing completeness applies to a model of computation, not to a physical instantiation of a machine. The stumbling block of "figure out how the brain works" applies more to the argument like the one I was responding to. How a person can know that a general model of computation can't implement the way people think, if we don't know how people think?

The existing LLM training methods on the other hand give the results that are hard to distinguish from "thinking like people," judging by the end results.

"They are fundamentally not suited to thinking like we do" stays wrong nevertheless. They are fundamentally suited to everything not proven to be outside their modelling ability.

> They’re fundamentally not suited to thinking like we do

LLMs with CoT are Turing-complete. So, theoretically, they can implement any kind of finitely describable algorithm (barring super-Turing computations).

LeCun calling them "world models" gives a high-level description of the desired functionality. They are Joint Embedding Predictive Architectures (with SIGReg). They might produce more useful world models, but it's yet to be seen.

> The last part is why they will remain illegal

Where are intermittent energy sources illegal?

We aren't dealing with individuals coming from intelligent species that have managed not to self-destruct (yet). AIs don't have the corresponding evolutionarily selected mechanisms. Those mechanisms need to be created de novo.

Provable correctness is much harder than focusing on the happy path. If a pipeline doesn't include a look-for-the-vulnerabilities stage to save costs, a model wouldn't go out of its way to do it. The models are trained to do what they are asked to do.

I forget to add the obvious: some garbage gets through despite the mitigations. In the limit of pure garbage input, you'll get a model that internalized garbage generation. And you'd be better off throwing it away and starting over.

This is simplistic to the point of being blatantly wrong. Training data isn't garbage. It's programs that do their job, but that are sprinkled with errors. Uncorrelated errors gets averaged out during autoregressive pretraining. Correlated errors can be somewhat suppressed during post-training. Hallucinations (of the generalization-error kind) can be dealt with using synthetic data that improves the model's generalization.

It seems to be just a quirk of fortune that it wasn't the other way around. That is that Europe wasn't decimated by American deceases.

Batteries compete with peaker plants for diurnal variance, but they can't replace peaker plants for seasonal variance. It's an interesting dynamic.

It reminds me of "At the time we drew boxes labeled 'perception', 'cognition' with arrows between them." An imprecise quote that I can't place.

I guess my box labelled 'subconsciousness' is trying to say that low-level mechanisms that give rise to the observed cognitive phenomena might have nothing to do with neat boxes.

Astronomically (or better to say combinatorically) large Markov chain can be used to describe a foundational model, but it doesn't capture generalization ability of the foundational model, which is demonstrated by post-training.

Programmers don't work reliably too.

_Radiative cooling_ is the only option because it's in vacuum. But you can transfer heat to the radiator either passively (using thermal conductivity), or actively (using coolant). The second option allows to effectively use larger radiators.

BTW, we seem to have Noisesphere instead of Noosphere.

They use passive cooling in their MVP satellite (for the sake of simplicity, probably). This, naturally, severely limits the heat rejection capacity.

Yep. There goes "hallucinations are inevitable" crowd.

I don't see why I shouldn't call people who choose wrong dangerous tools for the task morons.

The majority will not do it successfully. Rocket science is notoriously hard. Hopefully, there will be no incentives to go to such lengths.

Sending unguided rockets in the general direction of the enemy with 25% failure rate doesn't scale well to a 3-stage precision guided delivery vehicle.

Polluting LEO with debris is not the way to solve the problem of goods distribution. But, sure, there will be morons who consider it.

> the result of a brute-force attack

I would say "the result of easily scalable intelligent attack."

I've seen calculations for Denmark that has a great offshore wind power potential[1]. They indeed don't need nuclear. The lowest cost mix is offshore wind power (66%), combined cycle gas turbines (26%), and solar PV (8%). The costlier carbon-neutral option is offshore wind power (84%), solar PV (13%), and biomethane CCGT (3%).

Solar photovoltaic is so low thanks to its intermittency. The more solar you have as a percentage of the total production the more costly it gets. At the limit of 100% solar its SLCOE is about 4 times higher than nuclear (Fig. 3).

So, the availability and the quality (offshore wind is less variable, causes less backlash and so on) of wind power resources play a big role.

[1] https://www.sciencedirect.com/science/article/pii/S036054422...

I've seen "just slap a battery onto a steel mill to make it work with solar." So what? People on every side are saying dumb things. The required restructuring of the grid makes wind + storage + peaker plants + solar not so dominant as you picture it. Some countries can have it cheaper without nuclear, some can't.

Look at LFSCOE instead of LCOE. The electric grid can absorb a bit of intermittency relatively cheaply, but then you need to account for the system costs.

Note that the article has no info on whether it's FSD or Autopilot and whether they contributed to the fatal crashes. "Verified engaged" means ADAS was active at some point in the interval from T-30s to the end of the accident. The total number of collisions is not normalized by the miles driven.

I have a theory that they pushed the model away from human writing styles to not get copyright violation nags. In Opus 5.5 they've found a better point in the latent space of styles that is still far away from the human ones.

I didn't mean that as "we need hyper-intelligence." I mean that going all the way from nothing to singular, dual, plural grammatical numbers attached to particular nouns to abstract one, two, three, many to zero, one, two, three... to rationals to... is a very long way.

Replication of the development of mathematics by human civilization is a task for an artificial hyperintelligence.

Almost all of the latest models are MLLMs (multimodal large language models). For exmaple, [1] evaluates GPT-6 Astra on vision tasks.

[1] https://blog.roboflow.com/gpt-6-astra-vision/