Thank you! Let me wildly oversimplify and make sure I understand.

The fundamental problem is that if you train on a set that includes multiple independent facts, the generative aspect of the model - the ability to generate new text that is statistically consistent with the training data - requires remixing and combining tokens in a way that will inevitably result in factual errors.

Like, if your training data includes “all men are mortal” and “all lions are cats” then in order to generate new text it has to be “loose” enough to output “all men are cats”. Feedback and reinforcement can adjust the probabilities to a degree, but because the model is fundamentally about token probabilities and doesn’t have any other way of accounting for whether a statement is actually true, there’s no way to completely remove it. You can reinforce that “all cats are mortal” is a better answer, but you can’t train it that “all men are cats” is invalid.

To be fair, 2015 was definitely after he was a red flag, albeit for very different reasons than anything Saltman or musk care about

Yeah, all that “style over substance” nonsense is really strange given that those early sci-fi authors were more notable for cleverness and sheer volume of output than for consistent literary quality (and I say this as someone who also read and enjoyed a lot of Asimov and friends growing up). Like, Sturgeon may have coined the “90% of everything is crap” law, but when you write the amount that they did for the pulps you end up with some real gems in that 10%.

I thought we confirmed that his head did just do that, which is why the CIA had activated their sleeper agent in Lee Harvey Oswald to take a shot from the Texas schoolbook depository at just the right timing and angle to provide a mundane explanation that didn’t expose the flaws in their transdimensional mind chips.

In unrelated news my wife finally managed to get me started watching Fringe.

Eh. I can sympathize with the desire to provide up-to-date information while also wanting to CYA if anything changes or if you’re missing anything.

With a new context window, it responded as if the drift [in the previous conversation] had never happened.

Now, as I understand it this is literally the definition of a context window.

If nothing else they got the SCP wiki in there which gets into some of the noosphere stuff in the more esoteric and metatextual entries.

The metaphor I’ve used before is hammering a nail in with a shoe. It can work. If you have a lot of nail-hammering experience - especially hammering-shoe experience - you can find ways to improve how effectively it works. But by the time you’re able to use a shoe as anything resembling a hammer you should be able to both do the work better with the right tool, even if it is less convenient (needing to write the code yourself being analogous to needing to carry a big hammer with you) and more importantly recognize why it’s not an acceptable tool. Especially because in this analogy the only shoes are made of the finest orphan leather.

The problem is less that the system would somehow ignore that part of the prompt and more that “hallucinate” or “make stuff up” aren’t special subroutines that get called on demand when prompted by an idiot, they’re descriptive of what an LLM does all the time. It’s following statistical patterns in a matrix created by the training data and reinforcement processes. Theoretically if the people responsible for that training and reinforcement did their jobs well then those patterns should only include true statements but if it was that easy then you wouldn’t have [insert the entire intellectual history of the human species].

Even if you assume that the AI boosters are completely right and that the LLM inference process is directly analogous to how people think, does saying “don’t fuck up” actually make people less likely to fuck up? Like, the kind of errors you’re looking at here aren’t generated by some separate process. Someone who misremembers a fact doesn’t know they’ve misremembered until they get called out on the error either by someone else with a better memory or reality imposing the consequence of being wrong. Similarly the LLM isn’t doing anything special when it spits out bullshit.

Godspeed, @self. Take this as an opportunity to put it out of your mind and enjoy a well-deserved break.

Not that I know what to do with a break without internet access, but I’m told that our ancestors found ways to entertain themselves.

Man, I never would have guessed that “lmao spell ICUP” would have been one of the most valuable experiences I got from going to public middle school.

My first thought is to make a very unkind joke about his willingness to read when he could be watching.

We did catch it internally in testing (as we use VS Code for all our work, so some folks did stumble on it), but I think we underestimated the impact and should do a better job at that.

Either this is an outright lie or it’s a sign of just how fucked this industry has gotten. There should be no way that anyone looked at this and decided it wasn’t a big enough deal to block given that this is basically the single issue driving most of the industry’s cultural discourse and a good chunk of the broader world’s as well. If that’s what happened then the people making those decisions are so thoroughly insulated from literally any feedback that the industry - to say nothing of the world at large - would be better served if they were replaced by a literal magic 8 ball.

I don’t want to underplay how bad this is, but did BBC really need to use the “slutty anime witch” image of Ani for that story? Or was that the actual avatar he had set for it? Like, I’m not saying that it changes the problem or makes him less the victim here but it is yet another example of “goddamn why is this cyberpunk dystopia so cringe?”

I nearly bounced off when I couldn’t tell if his praise of Kissinger (spit) was ironic, but it was ultimately a very well-rounded examination.

They amplify what you tell them with no discretion despite their reassuring interface design. Thankfully I’m a genius who only has perfect thoughts to feed into it, so for me it’s an unambiguous positive.

I’m sorry, do you have Prediction Market Successes, us ove to rerhation mones?

What about Betting six norms of antiorizes on our beliefs?

You’re so non-empirical that you don’t even Graduates updating for rarhation of rationalists and advernces.

Off-topic, but the ongoing retraining process has hit a point where my wife and I are starting to throw out applications again after taking what ended up being a couple years off the market. Any tips or advice would be appreciated given that we’ve been out of the loop for a bit.

In particular, does anyone have advice on how to vibe-check smaller employers? My wife has an interview for an accounting clerk position and is concerned that she’s going to end up somewhere that practices one of the more hostile branches of Christianity or otherwise have an inevitable conflict of values.

Setting aside, for a moment, the flagrant racism and lack of historical and cultural awareness, the fact that the ships are mirrored across the center point because apparently the bow and stern of a sailing ship look similar enough to whatever model creates this image really does put this whole argument into context. Not that the people actually having those theological arguments appear to appreciate it.

A bus? You mean the Megapod?

Train? You mean the PodChain?