How does "zuzai uses zero AI" fix the organic vs pesticide free metric?
The only difference between it and "AI-Free," "no-AI," "non-AI," and "not by AI" is that you're using more words to say exactly the same thing. 'zai' -. "zero AI" would be precisely the same.
Author has pointed out an issue that the phrase carries just as much as "pesticide free" would in their own estimation, and simply jazz hands away the fact that the proposal is susceptible to exactly said critique. AND, I'd argue that both 'pesticide-free' and 'zero-AI' are better specifically because they draw attention to the issues caused by pesticides and AI.
There could be a difference between academic mathematical research and commercial mathematical research, just as in the difference between FOSS and commercial code. Perhaps the difference lie between pure math and applied math with applied math having more incentive for secrecy given that large swaths of computing is just applied math in fancy wrappers.
If you can derive some special formula that prints money from accurate prediction of asset valuations if and only if you're the only one with the efficiency, you're incentivized to keep it quiet.
If you're a nation state that's cracked a simple method to decrypt modern "pre-quantum" encryption schemes with less compute/wall time using a development in mathematics, it is in your interest to keep this quiet so that the parties you surveil to not switch or adapt. Similarly if you've figured out a handy new form of encryption, you're incentivized to prevent leaking of details as it is harder to reverse such operations without details of the methodology.
Or you've designed some special optimizer that you then bundle in a subscription service. Your business now only profits while the algorithm is secret. e.g. The DVD Copy Control Association's CSS encryption was only a valuable thing that the company could charge for or entice clients with before DeCSS was published/became popular.
The problem is wildly more complex than typically portrayed.
First of all, there is a lot of work already going in to slowing and eliminating aging. Every "please don't smoke" campaign is actually a "smoking causes early death so not smoking prolongs life" campaign, if you think about it. Healthy eating, cancer prevention and treatment, dementia etc treatment and prevention, brain training, physical and mental rehabilitation training -> all are efforts to reduce odds of untimely death meaning all are efforts to expand life expectancy.
Outside of this, consider the magical curing of age related death tomorrow. Do people with unbearable conditions now have to live with them forever? With their quadrapligia, lack of long/short memory, parkinsons, etc? Suppose we can magic hand-wave our way around those.
What of births? Approximately 130-140 million children are born globally each year. That's another billion in ~8 years. We already struggle to feed our current population with ~700million facing food insecurity, and estimates say that using all available space, not wasting, and being more efficient with our diets we could perhaps support 20 billion on this planet. This would, however, require perfect cooperation between everyone on earth deciding to share perfectly and agree to move to plant based diets where required. At current birth rate, we'd have less than 100 years to solve food waste and build out the capacity. Or is birth now controlled a la China? Who gets to decide? What happens when the decision turns out to be wrong and leads to inbalance of viable mating pairs? Do we pivot to global sterilization? No more births until we colonize Mars?
And where does everyone live? If we are maximizing use of arable lands, humans can then only live in un-arable lands because we need farms and greenhouses on those lands. So we have to live above the green-line on mountains, tundras and permanent ice, and deserts. And the more food we have to grow, the less land we have to live in, meaning density increases meaning that the odds of a super-spreader disease increases meaning that people will die regardless because in all of this we never cured death but only age-related death.
>And when I look at the problem... it seems entirely solvable!
If that were the case, that you really saw a clear path, you'd write the proposals and get the funding. If it is the case, rather, that what you have is simplistic hope/belief that it "should" be solveable, then you're in the same boat as everyone already doing the research, minus any actual research contribution.
There are a few of these "maintainer said no AI pulls, so we forked it" in the wild already and I am really interested to see how it all pans out.
Like, okay, sure, maintainer says AI bug monitoring is cumbersome. One fork stops reading the AI bug monitoring, the other leans in. Isn't the latter more likely to find/fix critical/performance/security issues?
>Similar to how noone would want to drive a car that was 100% hand built by humans
Though, of course, this highlights that the market will fragment itself given that there are in fact many humans who will outright refuse machine driven cars despite mounting evidence that they are vastly safer and more efficient in a growing number of situations.
We'll likely see both "individual who won't use AI assisted software gets hacked" and "individual who won't get in a robo-taxi kills jay-walker in broad daylight"
I find it really odd that unicode just ported all of these with zero context wrt their meaning and yet are still excluding certain things that "aren't in common use"
Assuming the AI is not run locally on my machine, how is the "privacy" handled?
Any PII stripping? E2EE with an AI instance on servers that "don't keep any logs"? Which model? Is it routed to some SOTA model, or routed at all, or all processed "locally" on Journal42 servers running local models?
I can't see any information at all on the site other than "trust us, we've got your back, spill your heart out, we promise to keep it quiet"
The protocol was only ever intended to be distributed, not anonymous. Simple logic will tell you that blindly connecting your raw IP to random other IPs is not an anonymous function, regardless of the protocol.
You can, of course, take steps to increase anonymity while using most any protocol.
>Will this make it easier for ISPs to penalize their irresponsible customers?
Only if they (the ISP) are so irresponsible as to have not noticed a decade ago that such monitoring was already possible.
I understand the "I don't have to worry about file compat or ownership (looking at you Sony)" aspect, but surely you've encountered at least one book whose contents you found to be disturbing?
As a counter-argument, an e-reader allows me to read many thousands of classics for the simple cost of the device and connectivity to Project Gutenberg etc. I can read thousands of years of history across tens of thousands of books for an ammortized cost of perhaps less than a penny a book.
To do the same, even relying entirely on a library, would cost me many thousands of hours in transit (and neither my time nor transportation are free) and waiting for inter-library loans. To do so in purchasing the books would cost me tens or hundreds of thousands of dollars.
Similarly, if I am studying a topic that requires access to more than one book, it is far easier to cross-reference and co-consume during longer commutes, or reading outside the home (especially camping/hiking) because the device ways the same regardless of number of books. I could easily carry perhaps 100k books in my pocket if I were so inclined.
Particular textbooks might be difficult though, I suppose, because I know that there is significant effort put in to making them less portable, no access to reasonable file formats, etc, and all of the "value add" in the linked videos, articles, and quizzes. But, I think those are more arguments against the text itself than the medium, unless you're taking a course that requires a specific book that is particularly anti-portable.
>I have a feeling that being transgressive is universal
It certainly must be, no? Why would we have to culturally encode taboos and legally enforce particular behavior if not for a persistant presence of transgressive behavior?
No one would have sat down to write: "Any person may possess any number of live frogs to use in frog-jumping contests, but if such a frog dies or is killed, it must be destroyed as soon as possible, and may not be eaten or otherwise used for any purpose." (CA Fish & Game Code ยง 6883) unless there were in fact people eating their war-frogs AND people who believed that said war-frogs deserved proper funeral rites.
Would be nice if desktop rendered in columns. Would be nice if it were curl-able (like not serving raw html to curl). Would be nice if URL pattern was like wttr.in (NYC for example is https://brolly.sh/forecast/YcUDdBLc <---which means nothing to me, vs wttr.in/nyc <---which is plainly obvious)
I'm not a fan of how many "I set budget X and woke up to an eleventy trillion dollar bill" posts I see, and those are all generated by companies giving their tools the 'judgement' that the completion of the task is more important than the users wallet (or more cynically that they think they can actually get paid by just blasting unintended compute)
Clicked thinking "well, duh, because language has only been independently invented like 4 times and the vast majority of languages are merely derivitive, and any English speaker is exposed to a majority of these languages because English has tons of loan words"
But was pleasently surprised with a neat language learning tool.
Is it really all that different from how the models "know" anything else?
The best way to know that christmas trees are typically purchased in december isn't to access historical financial tables and run the entire analysis yourself but to just "know it" from the articles you read attesting to the fact.
If you've found data that is the answer to the question, be it a random user question or a benchmark, the softwares' goal is to produce the correct solution and it is cheaper to retrieve from storage than to compute.
It's the same with coding. The agent usually isn't really thinking about the problem from first principles, it's just giving you the answers it's already found in instances where someone else asked the same question.
It seems to me that if you're really testing for reasoning capability, just as if you were an instructor administering a test, you'll need to change the test from run to run in order to make sure the agent/student isn't just copying old tests.
The only difference between it and "AI-Free," "no-AI," "non-AI," and "not by AI" is that you're using more words to say exactly the same thing. 'zai' -. "zero AI" would be precisely the same.
Author has pointed out an issue that the phrase carries just as much as "pesticide free" would in their own estimation, and simply jazz hands away the fact that the proposal is susceptible to exactly said critique. AND, I'd argue that both 'pesticide-free' and 'zero-AI' are better specifically because they draw attention to the issues caused by pesticides and AI.