rocqua

I believe there might be more to it. Wasn't a big part of thinking or reasoning taking the response, replacing the final period with "Wait!" and then continuing? Which suggests that such words actually are important to the internals.

TLDR one developer tried to trademark the name meschcore. He also took over development of some parts and then hid that he was vibecoding.

Since being called out on these things, he’s been ghosting the rest of the team. He also happens to control the former official site.

This delay in nuclear means that, by now, nuclear is no longer a solution. Had we started with nuclear somewhere in the first decade of this millennium, then it would have been a good solution to climate change, and we would be in a better position.

But in our current position, nuclear is too slow, and luckily we have alternatives in wind and especially solar. Where the main advantage of solar is how quickly it’s scaling. Notably, the slowness in building nuclear also limits how fast you can improve the process of building nuclear. Whereas solar is so quick to build you can learn lessons and try innovation much faster.

I hate to be this dismissive, but it feels like an academic with a paternalistic streak looked deeply at how the Internet works, saw lots of different protocols and weird design decisions, and decided: this is not coherent enough. Then he figured, I'll make all the decisions now, that way it'll be coherent. And let's give every subnet a centralised source of trust and management. That'll make the design so much cleaner!

By which I mean to insinuate there's a lot of nuance and learned lessons in the current situation that this design seems not to learn from. Even though it did learn some lessons, I don't think this passes 'Chestertons fence'

There was a lot of forceful diplomacy by the US. Sure, but there was also a lot of actually good diplomacy happening. Calling all of that a thin glove is underselling the good work of a lot people.

The good side of US diplomacy was one of the most positive forces in the world. Trump fully dismantled that. Not just the US aid work, but also the Pax Americana that really limited the scale of war in the world.

There were horrible missteps at the same time. The US wasn’t all good. Maybe it wasn’t even net good. But there was a significant good side, and its dismantling isn’t a small thing in the world.

Confidentiality of the TLS connection is indeed easy to handle here.

The hard part is certificate authentication. And that's not included in the cipher suite setting.

I've heard it a lot from podcasts that are towards the abundance movement. I think its common within the rationalise movement.

Personally I really like it for "load-bearing assumptions". Because it let's you work with assumptions whilst pointing out the potential issues of that assumption.

Nonono. Nothing mandatory. It's a parents choice. The thing to do is help parents enforce their choices. Not forcing choices onto parents.

It's not that they are meaningfully different. It's just acknowledging if you really want currying, you can say 'why not just use a single parameter of tuple type'.

Then there's an implication of 'sure, but that doesn't actually help much if it's not standar' and then it's not addressed further.

Someone else in the comments mentioned that scala does this with _ as the placeholder.

For schools and hospitals, why specify COTS? Do you want SOTS for schools and HOTS for hospitals just like we have MOTS?

Figure out where you can't buy pictures to narrow it down, if you want a more exact match, pay for pictures from that area from non US providers.

If the solution is parental control software, that also puts onus on operating systems to present the means for such software to work properly. This does not mean the OS should censor, it might mean the OS offers a censorship interface.

At least we seem to agree the solution lies with better tools for parents.

So the solution is effective parental controls. Government mandated age verification isn't parental control, and is unlikely to be very effective.

That means making it possible for parents to actively block bad websites, and making that hard to circumvent.

I think intentionally and willingly doing something whilst informed of the consequence doesn't count as human error. At least not in this context.

Though it would make more sense, since these humans are likely largely erroneous.

There's likely natural experiments in cases where police was misinformed either way about the danger of the suspects being arrested.

Age limits might be an alternative. Say at 65 or 70.

That's at an age where wizened legislators can move into advisory roles, instead of needing to find a next career.

When you put an LLM in reasoning mode, it will approximately have a conversation with itself. This mimics an inner monologue.

That conversation is held in text, not in any internal representation. That text is called the reasoning trace. You can then analyse that trace.

If you want to be pedantic about it you could phrase it as follows.

When the LLM was in reasoning mode, in the reasoning context it often expressed statement X. Given that, and the relevance of statement X to the taken action. It seems likely that the presence of statement X in the context contributed to this action. Besides, the presence of statement X in the reasoning likely means that given the previous context embeddings of X are close to the context.

Hence we think that the action was taken due to statement X.

And that output could have come from an LLM introspecting it's own reasoning.

I don't think that phrasing things so pedanticaly is worth the extra precision though. Especially not for the statement that inspecting the reasoning logs of sn LLM can help give insight on why an LLM acted a certain way.

This seems great in concept, and totally infeasible. But if anyone can do it, unicode seems like a great candidate.

Does anyone have reason for more optimism?

I recall similar advice around mylar heat blankets. Perhaps those got mixed up?

Feels quite power-law like to me, but checking this roughly it seems to decay more quickly than a power law, but with a fatter tail. At least in the top 1000

The top 3 with 4,000,000 words have about 20 times as many words that the 0.14% percentile (at rank 1000) with 200,000 words.

In between (at rank 500) your at about 450,000 words, so its not a true power law. Because a drop of a factor 9 per 500 ranks would suggest that rank 1000 were at about 50000 words.

So for the ISS at 20c you'd get 481 W/m^2 so you'd only need 2.3m2. So comparing the ISS at 20c to space datacenters at 70c you get an improvement of 63%. Nice, but doesn't feel game-changing.

The power radiated is T^4, but 70c is only about 17.1% warmer than 20c because you need to compare in kelvin.

Why does it matter if people have to sell their shares to unlock value? Is it just the friction of small orders?

Buybacks for manipulating share prices and earnings per share are indeed silly. But they should also be trivial to compensate for by normalising on market cap instead of a single share.

A buyback is almost the same as a dividend, with minor differences around tax and effects on derivative pricing.

And ASML has been paying out a dividend for a long time.

Just say "we are storing your keys on our servers so you won't lose them" and follow that with either "do you trust us" or even "we will share this key with law enforcement if compelled". Would be fine. Let people make these decisions.

Besides, bit ocker keys are really quite hard to lose.

I find that the LLMs are good at the 'glue code'. The "here's a rather simple CRUD like program, please tie all of the important things together in the right way". That was always a rather important and challenging bit of work, so having LLMs take it of our hands is valuable.

But for the code where the hard part isn't making things designed separately work together, but getting the actual algorithm right. That's where I find LLMs still really fail. Finding that trick to take your approach from quadratic to N log N, or even just understanding what you mean after you found the trick yourself. I've had little luck there with LLMs.

I think this is mostly great, because its the hard stuff that I have always found fun. Properly architecting these CRUD apps, and learning which out of the infinite set of ways to do this are better, was fun as a matter of craftsmanship. But that hits at a different level from implementing a cool new algorithm.

So he explains OFDM in a way that implicitly does Amplitude shift keying.

I guess if you want to use different modulations you treat the complex number corresponding to the subcarrier as an IQ point in quadrature. So you take the same symbols, but read them off in the frequency domain instead of the time domain.

And I guess this works out quite equivalently to normally modulating these symbols at properly offset frequencies (just by the superposition principle)

If your value proposition is: do X, and then you have to take action against an open source competitor for doing X better, that shows that you were beaten at the thing you tried very hard at, by people with way fewer resources.

I can see why you would call that embarrassing.

Sorry to hijavk, but why are drones still on electric power? I'd expect some kind of air breathing engine to be manufacturable at low cost if it only has to function for a single flight. Especially for interceptor drones the speed is valuable. But for all types of drones the extra range is also highly valuable.