twothreeone

Cough we're still waiting for those alignment dollars to really kick-in!!1

> It's worth noting that not everything that's good for you is enjoyable, especially in the moment.

100%.. people tend to understand this much better with physical exercise.

Social connections have as much of an impact on mortality as e.g. smoking [1].

Yes, quitting smoking, working out, and talking to people is not pleasant - it is also healthy and good for you.

[1] https://pmc.ncbi.nlm.nih.gov/articles/PMC2910600/

That doesn't mean you can hit it too. The only case which I remember was well documented was for a turret gun "controlled by satellite" hit the passenger of a standing vehicle from the back of a truck in front of it (nuclear scientist Mohsen Fakhrizadeh, killed in Tehran in 2020 while his wife, sitting immediately beside him, was unharmed). Also those reports were pretty questionable at the time, there were multiple competing narratives, and I'm not sure the satellite gun story was ever independently verified.

Same!

I'm pretty sure that is a mis-characterization of what actually happened: a 3rd party app collected user profile data from FB _through somebody's friend connection_ [1].

The Zuck then apologized for this undocumented API access leak and FB subsequently patched it out of their platform. It had nothing to do with running ads. We shouldn't let time paper over the details here.

[1] "It also gathered data from their Facebook friends, which reportedly resulted in Kogan having access to the data of millions of Facebook profiles". https://www.cnbc.com/2018/04/10/facebook-cambridge-analytica...

Worthless is a pretty good description IMO in the context of what Lean is trying to achieve: "enable correct, maintainable, and formally verified code". Tens of millions of lines of LLM vomit may be many things, but it often turns out to not be correct and certainly not maintainable. Formally verified remains as a thin fig leaf covering the uncomfortable truth that formal methods only provide assurances under assumptions (your toolchain, libraries, compiler, OS, and hardware are "correct" and don't expose some exploitable flaw).

It doesn't mean that it cannot improve over time, maybe the proof can be "minified" to a state where human reviewers are able to comprehend it; but as it stands there isn't really much insight or confidence to be gained from the artifact itself.

Right, exactly that was what I meant. Thanks for clarifying. I remember being surprised by this when I heard it on NPR a few weeks ago.

I mean they're right, rising oceans will be a drop in the bucket for the majority of us. Unless we act together, human civilization will end through the combination of breakdown of global trade, mass migration due to heat and resource wars, and collapse of our food chain. Ocean levels are just one of the side effects that goes along with rising global temperatures.

Beef production in the US is already down from 30 years ago. At some point you won't be able to raise cows anywhere, less alone feed them. Corn and wheat have seen tremendous fluctuations in the past and will likely continue to fall. Imagine corn yields dropping by more than 80% in the US [1]. Of course MAGA people will probably only notice once the shelves aren't being restocked.

[1] https://www.pnas.org/doi/full/10.1073/pnas.0906865106

The implicit "men born outside the US are not men" is arguable worse.

In the first case they (the US govmnt) could hold him (the citizen) in contempt (in a cell) indefinitely.

While I tend to agree on the overall sentiment, I think this rebuke is inaccurate. Some of these "reasoning" models are trained using "Chain-of-Thought" where the model is presented explicit, intermediate reasoning steps (either by a human or some automation) that supposedly get it closer to the correct answer. These intermediate steps are what was originally called "thinking traces" - not what the model produces to mimic them.

But yes, anthropomorphizing model outputs leads to worse outcomes.

Well, one obvious question is if the simulated brain experiences pain, or more generally can it suffer? And if so, how would we tell.

But Matthew, think of the kids!!1

Honestly though, framing this as a "tech issue" doesn't help IMHO, it just muddies the water. Ever since RSA was invented privacy has been about educating people on how to use it effectively and _why they should care_. If voters now are choosing authoritarianism over democracy and individual freedom, I think we have to face the reality that after almost 50 years of fighting battle after battle on the technology front, we've largely lost the war on the home front in this regard.

Yeah. This is basically why we need to "accelerate"..

There are 189 active legislative bodies to choose from; it's hard to imagine, but yes in the worst case they all fall in lockstep on this issue and none of them provide this kind of freedom. The obvious next step is to buy a remote island (better yet: some foreign body in Space) and form "Freedom (tm) Republic".

So presumably the next step is they move their servers out of the US and then DA's target the Linux Foundation directly (e.g., coerce the kernel to take a change by some Meta engineers that "attests" a device when said option is compiled in, then force it as default on.. so in Debian you'd get an "Illinois"-kernel variant).

At which point the only move is to relocate the foundation to another country, which will then of course be hit by tariffs and sanctions.

TL;DR: They claim it's Meta (which yeah it would make a lot of sense).

Chris usually takes an educational angle, I don't think this is LLM-generated content at all it's just his style. I highly encourage watching some of his DefCon or BlackHat talks, they're fun!

Yes, plant more trees!

> I’d also argue that the vast majority of their capex exists to support these two load-bearing failsons.

Gotcha.

I agree it feels scary from the employee's perspective, but at least in this case, it doesn't seem like it would've been difficult to avoid: simply return your work devices when handing in your badge:

> [...] multiple former Apple employees now working at OpenAI reached out to discuss returning Apple-issued work devices they kept when they left Apple.

By not doing that you effectively just cry "sue me", this is stupidity on the employee's part (or malice). Oh and maybe don't take "screenshots of confidential documents".

I agree, I actually saw specks when viewing the Odyssee on 70mm IMAX. Sorry, but how is that "better" than any clean, digital projection? And the hoops they apparently jump through to make it work during the shoot to get to that point. Pure nostalgia hysteria.

Eventually they make you fill out some paper work, or pick up some boxes from somewhere and bring them to another place. You know quick human tasks, just to verify you're really human. It's just a verification check, nothing else.

Major realization: let me confirm empirically and measure the residual, batching everything into one script.

> To prove that Ω is computationally and therefore logically irreducible, requires a theory of program-size complexity that I call algorithmic infor- mation theory (AIT) [Chaitin, 2005]

Interesting, I think everyone else calls this Kolmogorov complexity.

Wow, very cool "The European way to pay online." - so surely they are on f-droid or have an apk download like Signal with build signatures published on their website! Oh, no - they require you to have a Google account and Play Store on your mobile device. Well, false advertising again. Disappointing, but why am I not surprised. Thank you, but no.

If this is true it almost sounds like DeepSeek is following the Anthropic playbook of trying to pressure the local government into aligning with their corporate agenda through scare tactics. So I wouldn't be surprised if Liang Wenfeng "disappears" for a little while from the public eye in a few weeks.

It starts at page 148.

Just fund your own group of grave diggers with PhDs and an attached museum.

I think it's similar with other liability issues, e.g., when a company happens to "lose" customer data through a breach. They will be on the hook for not having certain audits and certifications at regular intervals. Practically never will anyone be feeling any pain due to absolute disregard for basic common sense precautions to prevent issues in the first place. So usually, the pattern is that issues that can be outsourced to insurance will be handled by compliance departments - which don't care about the actual problems, just that the fallout from them is "managed" accordingly.