Around the time this website was made, I was building an application for a big company in Spain that was to run as a Java applet and required the code to be signed.
They did not yet have their own certificates so I had to make my own CA during testing and sign the code, and I wanted to make sure that they did not forget to switch to their certificates later, so instead of signing the code with my name which some bureaucrat might decide to not bother changing, the code was signed by Britney Spears.
They noticed it, got the joke and made sure to switch certificates for the release. Everything went well thanks to Britney.
It seems like they ought to have at least clarified the context, though. "Large integer precision error in Bash command output rendering" doesn't mean anything to anyone. If anything, it sounds like a bug in Bash.
It's also not a real rule. If you use the original title, they will regularly "fix" "misleading" or "clickbait" titles too. All of which is completely subjective at the whim of the mod team with zero transparency.
A particularly interesting part that I did not expect from the title:
> Before the rats encountered the detour, the research team observed that their brains were already firing in patterns that seemed to "imagine" alternate unfamiliar mental routes while they slept. When the researchers compared these sleep patterns to the neural activity during the actual detour, some of them matched.
> “What was surprising was that the rats' brains were already prepared for this novel detour before they ever encountered it,”
Suppose we simplify the scenario and think of experiences as draws from a discrete probability distribution, e.g. p=[0.1, 0.1, 0.7, 0.1].
Suppose further that all events are a draw of type 1, 2, 3, or 4, and that our memory kept a count and updated the distribution - it is essentially a frequency distribution.
When we encounter a stimulus, we have to (1) recognize it and (2) assign a reward valence to it. If we only ever observed '3', the distribution would become very peaked. Correspondingly, this suggests that we would recognize '3' events faster and be better at assigning a reward valence to those events.
Then if we ever encounter a non-3 event, we would recognize it more slowly - it is well-established that recognition is tied to encounter frequency - and do a poorer job assigning reward valence to it. Together this means that we would do a bad job selecting the appropriate response.
Perhaps this scenario-based dreaming keeps us (and rats) primed so we're not flat-footed in new scenarios.
The question then becomes - if these scenarios are purely imagined, where are they being sampled from? If we never observe 1, 2, and 4...how do we know that these are the true list of alternative scenarios?
Yeah this part was pretty weird. How do they know that was caused due to the rats' brains firing because they were imagining unfamiliar routes, vs something completely unrelated to the maze routes at all? Just because the hippocampus flash patterns matched doesn't mean that's what the rats were thinking about while sleeping I'd think
> Overall, I don’t think the LLM saved me any time. I didn’t end up keeping the code it gave me. But it got me unstuck, and that meant I actually made progress.
I haven't tried LLMs yet but this sounds valuable.
In 1984 one of the developers of Xanadu at the time, Chip Morningstar, wrote a technical description to apply for funding.
In typical Xanadu fashion it was kept secret, protected by Dark Magick:
> He who transgresses against the propriety of the Information contained herein shall be Cursed! Woe unto all who reveal the Secrets contained herein for they shall be Hunted unto the Ends of the Universe. They shall be afflicted unto the Tenth Generation with Lawyers. Their Corporate Bodies shall be Broken and cast into the Pit. Their Corporate Veil shall be Pierced, and Liability shall attach to the Malefactors in personem. They shall suffer Ulcers and Migraines and Agonies Unimagined. Yea, Verily, for such shall come to pass against all who would Dare to Test the Powers of Xanadu unto their Doom.
It was believed to have been lost forever but in 2019 he found the only known remaining copy, a printout he had at home buried in old stuff, and posted it with OCR in his blog. He wrote:
> At the time, we regarded all the internal details of how Xanadu worked as deep and dark trade secrets, mostly because in that pre- open source era we were stupid about intellectual property. As a consequence of this foolish secretive stance, it was never widely circulated and subsequently disappeared into the archives, apparently lost for all time. Until today!
I took it and made a cleaned-up augmented HTML edition, with additions like Engelbart’s structural statement numbers which allow linking to individual paragraphs:
It includes a glossary of terms and detailed explanation of the content addressing “tumblers”, and is the best technical description of Xanadu that I know of.
That man, Rex Malik, participated in (among other things) the 1982 BBC series “The Computer Programme” (https://en.wikipedia.org/wiki/The_Computer_Programme), typically in a small section at the end on an episode but also as narrator in other parts and is credited as “Programme Adviser”:
Edit to add: I see now that you also did the Uprintf submission from a few days ago (https://news.ycombinator.com/item?id=44960664) which had some baffling bugs, and then you mentioned that you were using AI for some of it. That might explain the missing parts here too.
This is possible not only fault with AI, because I always do tests, but sometimes I don't always manage to cover all platforms like this Linux, because I have to admit that I don't have a PC, and I did all this on my phone.
The bottom-right camera view of Jonathan Blow and its audio do not match exactly the timing of the screen capture that fills the image. There are several seconds of delay.
It can be seen clearly at this point: https://youtu.be/RIYGaSBKy3w?t=1652
The most obvious moment is when he raises his hands gesticulating but the screen capture shows him moving the cursor in Emacs.
That confused me in some earlier points where he was mentioning things that were not visible on the screen until later.
Still, the talk is worth watching. I am particularly interested in the use of metaprogramming to ease debugging by gaining visibility on what the program is doing.
Edit: the delay has been reported in the YouTube comments and they say they will fix it and re-upload.
They did not yet have their own certificates so I had to make my own CA during testing and sign the code, and I wanted to make sure that they did not forget to switch to their certificates later, so instead of signing the code with my name which some bureaucrat might decide to not bother changing, the code was signed by Britney Spears.
They noticed it, got the joke and made sure to switch certificates for the release. Everything went well thanks to Britney.