6177c40f

I read that initially as "Forthtran" and I can't tell if that sounds like a good idea or an awful one

> "manufactured by robots from local resources or asteroid mining."

I feel like this just creates the new problems of operating, powering, and maintaining the robots, plus all the difficulties associated with mining asteroids and refining the mined materials somewhere (presumably somewhere that's not Earth's surface, which means we still need to generate power somewhere off Earth for this to work).

This seems like a chicken-and-egg problem. Wouldn't you need power in the first place (and therefore transmission lines) to even extract and refine aluminium locally?

If you want some more things that ChatGPT isn't good at, try to get it to form novel anagrams, palindromes, or other such wordplay. It's good at regurgitating textbook examples of those, but I found you can trip it up by asking it to do things like "Create a palindrome containing the word 'coffee'"

Depends on the kind of math, it can do basic things reasonably consistently, but for other things it will confidently respond with the wrong answer (e.g. matrix arithmetic). In my testing, I found that it will absolutely fail every time at tasks like "sum the digits of the number 1283892"

This particular example doesn't really alarm me, but it does make me think about a scenario that I believe is plausible in the near future. Imagine a world where AI enables anyone to write complex programs, regardless of skill level. In this case, we suddenly have a world where essentially anyone with malicious intent can lay bare vulnerabilities, even quite subtle ones, in critical software infrastructure and then generate code to exploit them- with little-to-no tech expertise required.

It seems that society becoming dependent on software was conditioned on the assumption that very few have the ability to exploit it and that fewer still have the ability and are malicious. But that assumption may not hold very soon.

However, there may be ways that AI can be used to improve security and robustness of software infrastructure to attack as well, so it may not be so bad. Time will tell.

There was a debate about whether it really couldn't be better explained by something other than life, then there was a debate as to whether they actually detected phosphine in the first place. See, for example, https://www.washington.edu/news/2021/01/27/phosphine-venus-s...

I feel like spectroscopy would only give you suggestive evidence of life at best, like if significant quantities of oxygen were detected in the atmosphere of an Earth-like exoplanet in the habitable zone. I think your best bet for direct detection of life would come from Mars rover or maybe a mission to one of the Jovian moons (but AFAIK no probe will arrive at a Jovian moon in this decade). Not sure either will detect life in this decade, but I sincerely hope you win your bet!

Isn't WASP-39b a gas giant? Is it even possible for such a planet to have active volcanism?

Esoteric programming languages. When you remove the need for a programming language to be useful or understandable, what you get is a combination of challenging puzzles, humor, and some actually interesting ideas.

Right now I'm actually working on a new esolang that combines an almost-reasonable stack-based programming language with Malbolge-inspired (https://en.wikipedia.org/wiki/Malbolge) obfuscation.

How about you give it a try yourself and post the results to HN? I would be curious what you find.

That this is being supressed is a bold assertion. Doesn't this seem like an odd, arbitrary thing to suppress? In situations like these, it's been my experience that the more likely explanation is that the evidence for it is a lot more spotty than you and the authors of the articles you linked seem to think.

Same here, in fact I'm rather surprised to learn that this isn't the norm for people with ADHD, apparently.

On a related note, my office is currently doing a sort of ad-hoc, informal hybrid work schedule, so some days of the week there's very few people in the office. On those days, I find I'm so much more productive just due to the fact that there's less commotion and fewer people talking. WFH was great for my productivity as well, of course. Anyone else notice the same?

I highly doubt that, even if everyone did this religiously, it would have any measurable impact on emissions, given how emissions from residential energy usage make up only a relatively small fraction of all emissions (and emissions from cooking only a small fraction of that). Am I wrong?

Thanks for linking this, I love this kind of "computer science fiction" stuff

They're worried about the case where the AI figures out how to get others to (unknowingly) pay its energy bill for it, in a manner of speaking.

What about it makes it not science?

In fact, the genus Cinnamomum which contains cassia and ceylon (C. cassia and C. verum, respectively) is pretty large and diverse. Sometimes what people refer to as cassia can actually be one of several species, e.g. C. burmannii, which is also often referred to as korintje. There are others that can be ambiguously named. E.g. the aromatic compound camphor is harvested from camphorwood (Cinnamomum camphora), which is not to be confused with the tree known as camphorwood in Australia, Cinnamomum oliveri.

I have no real point here other than cinnamon (i.e. the genus Cinnamomum) is fascinating and confusing in equal measures and people should definitely read up on it.

Yeah, I would argue that calling it counterfeit is completely wrong. Cassia and Ceylon cinnamon are both in the genus cinnamomum, so it's completely appropriate to call both cinnamon. Really, they're two very different spices whose flavors are only somewhat similar and have different culinary uses, but Ceylon cinnamon is less common in the US, so when we say cinnamon we pretty much always mean it to be cassia cinnamon- no deception or counterfeiting involved here.

I'm not really sure what they're trying to accomplish here, but I never thought I'd read a quantum physics-related paper with the sentences:

"We place a tardigrade tun on a superconducting transmon qubit and observe coupling between the qubit and the tardigrade tun via a shift in the resonance frequency of the new qubit-tardigrade system. This joint qubit-tardigrade system is then entangled with a second superconducting qubit."

Fairly accurate, though I love CSAs (and I suspect this is a loving parody by the author).

What I find especially funny is

> What the hell am I supposed to do with all these turnips?

Do people really not know what to do with turnips? Turnips are amazing- roast them, put them in soups, they're great. Now radishes, on the other hand, those are a mystery to me.

I've been thinking something along these lines recently too. A single neuron seems like an absurdly complex device. I doubt they are really similar at all to "artificial neurons" except in some vague abstract sense. So the brain is essentially a network of billions of powerful, hyper-efficient computers (not to mention all the other nerves in the body), whose function, communication protocols, and network topology have been developed for millions of years in such a way that has lead to human intelligence.

And yet it still takes decades to develop a human mind. Even just to become competent with a single language takes years and it requires much more than just learning that language. If we could develop an AI system that hypothetically could achieve human intelligence, could we even train it to that level faster than it takes a human mind to mature? Is there some kind of maximum rate at which the process by which "intelligence"- whatever that may mean exactly- can be reliably run?

And let's not forget that many of them are probably browsing HN at this very moment. They could be you, they could be me!

Would you not be saving (or at least not spending more) money by not having to pay for your healthcare needs. Besides the fact that if the rents really are cheap, then that would also be a huge savings for a lot of people.

This runs counter to all of my recent working experiences. What sort of things have you experienced that have led you to this point of view?

I had been messing around with a similar concept not too long ago in OCaml- as it turns out, strong, static typing can be very nice to have in writing ML models (and code in general). I find it reduces some of the cognitive overhead that comes from having to keep track of how the functions interact in the program (or perhaps, it reduces/clarifies the potential ways the functions can interact).

I'll be very interested to see where this project goes.