317070

> fight corruption and establish meritocracy, accountability, and the rule of law

Keep into account that the baseline for governing Serbia is low. As long as they can lower the corruption, they can do a lot wrong and still come out ahead.

Specu-less would be the English etymological equivalent. Like careless or tasteless.

They probably can tune for it, but it is not what the shareholders want anyway. I am unaware of any frontier lab who is actually interested in building something like that, long form proza indistinguishable from humans. Automated code and low hallucination rates are more important topics, and trade-offs need to be made.

Edit: why the style is like marketing corp-speak has to do with personalization not working well yet. So the default vanilla style is very corp. Most chatbots allow you to change the system prompt though and that can get rid of a lot of these issues in a lot of chatbots. But it is hard to maintain such a system prompt as the bots change so quickly.

Language models don't have a characteristic style you can spot from a mile away. They haven't had it since before ChatGPT.

What is true, is that chat models (and by extension agents) have a very characteristic style after their LLM went through post-training. But the pre-trained language model doesn't have that style.

So if you want to write well with an LLM, you could and it is a well understood problem. It is more a matter of that not being of much interest, as it competes with having the model be good for chat interfaces.

How to have a well-aligned, RL'ed, safe or thinking model, which also writes well in long-form, is a bit an open problem. Though presumably agents could do it when they have access to a pretrained model.

Not OP, but the message has a curious mix of ASCII with non-ASCII characters, like the same apostrophe character is sometimes in ASCII, and sometimes not. There is even a rogue em-dash in there.

I moved to the UK, and also got a dryer at that point.

The weather in the UK can be very humid, and if you don't have much of a garden, it can be impossible to dry your clothes at times. We've had times where our living room was full of clothes for more than a few days, and it wouldn't matter because it wouldn't really dry anyway.

So we've invested in dehumidifiers and a dryer, and it made our life a lot more comfortable.

If a fix existed, then why didn't Danfoss have a fix?

I agree that they had a firmware update, but unless they had the operational capacity to update the checks notes freezers of all their customers, that is not a fix. The theoretical and technological aspects of solving a problem are often the easy parts.

It's like if NASA would claim to have a fix for the Opportunity rover on mars to get it operational again. All they need to do is to replace the batteries. Well, that is not a fix if there is no way to actually replace the batteries in practice.

I don't see why the existence of a firmware update would remove the culpability of negligence from the hardware manufacturer.

> but I don't see how

LLM's are trained on human knowledge and taste. They are actually pretty good at deciding if a conjecture would be found "interesting" by the mathematical community or not.

Note that I am saying LLM, and not chatbot or agent. But even a chatbot can often still reasonably rank a list of mathematical statements by vague properties like "interestingness".

How to RL this is a bit of an open question, but there are interesting conjectures of how to do it.

All reagents are symmetric, so how do you design a reaction where you produce more left or right? So starting from purely symmetric reagents and catalysts, how do you produce asymmetry?

When the chirality of the product is an autocatalyst to its own chirality. So if your reaction starts off with a little more left than right due to molecular randomness, the left is an autocatalyst which means there is a runaway process and you end up at the end of the reaction with a lot more left than right.

If you repeat the reaction from scratch, it is random if you end up with a left product or a right product. But the product of your reaction is not a 50-50 mix left-right! I.e. you created asymmetry in your solution from purely symmetric solvents.

I have seen these parodies before, and while they are accurate in style, they seem to misunderstand or misrepresent Curtis' message. A message which is not that hard to understand, except that it falls comfortably outside of the Overton window:

"Driven by the seductive post-war fantasy that human beings and nature are self-regulating machines -- a cybernetic delusion shared equally by California tech-utopians, holistic ecologists, and free-market economists -- we willingly surrendered political power and moral responsibility to automated, network-driven systems, only to find ourselves trapped in a static, dehumanizing loop that cannot envision an alternative future and leaves the existing, unaccountable hierarchies of power completely intact."

But if you apply the logic that led to the conclusion of "natural equilibrium", truth of the matter is that we wouldn't be able to say if the waiting will be over before or after the heat death of the universe.

Even the "million years" is an overoptimistic view, and probably reflects the idea that the ecosystem was in some kind of messianic state of equilibrium before man arrived, which happened roughly in the last million years. But even that is broadly a fiction. Ecology is not a system, and it is not stable even without human intervention.

It's because the models have different purposes in decision making, in ways which decision makers don't often appreciate.

Models are still great at helping us understand the world and are in many ways the best thing we have. The problem is that today we are overly relying on them to make real policy interventions on an ecology (e.g. what is a "healthy" amount of animals to cull or fish) based on an ecosystem, which is just a poor model.

If we could get rid of this idea of "stability" and "equilibrium" in economics and ecology, I would be a happy man.

> Ironically, the experts, the ones building the "wrong" models, tend to be the ones most aware of this

Exactly why I'm here :-)