What's interesting is that the LLMs' coding personalities seem to match their policy WRT to strategy, which suggests an underlying consistency.
Claude, for example, is very eager to begin coding, and very persistent. It tends to exit plan mode even when the plan is half-baked, and will go as far as deleting tests to get the suite to "pass."
ChatGPT on the other hand is very hesitant. It loves to pause and ask for permission before it starts coding, and gives up quickly if it runs into a problem. This is similar to its tendency toward passivity in the strategy simulation presented here.
The article is really vague and a bit misleading, but the bill text appears to be surprisingly readable, and honestly not much longer than the article.
A lot of agricultural water usage (more water than AI) is for growing corn to turn into ethanol so we can add it to gasoline. It's not a small amount either, 40% of all corn in the US is used for this purpose.
This was the same logic that was used when building nuclear weapons, and many of the scientists involved in that tried to find a different path (most notably Niels Bohr). I think we would be in a much better world if they had been successful. It's good that we're trying again w/ LLMs.
They also removed the words build, develop, deploy, and technology, indicating that they're no longer a tech company and don't make products anymore. Wonder what they're all gonna do now?
Could also be great for maintenance dosing. I'm reaching the end of my ~weight loss journey~, and it's not a sure thing that the insurance company will continue paying for the injections once I'm no longer overweight.
I'm definitely willing to keep taking it. If insurance won't pay for it, I could pay for the pill out of pocket if I had to, which would be cost prohibitive for the injections.
You can't call something a bad-faith argument just because you disagree with it. I mean, you can, but it's not at all convincing.
As I said, if AI companies reproduce copyrighted works, they should be sued, just like a human artist would be. I haven't experienced that in my interactions with LLMs, but I've never really tried to achieve that result either. I don't really pirate anymore, but torrents are a much easier and cheaper way to do copyright infringement than using an AI tool.
Every artist and creator of anything learned by engaging with other people's work. I see training AI as basically the same thing. Instead of training an organic mind, it's just training a neural network. If it reproduces works that are too similar to the original, that's obviously an issue, but that's the same as human artists.
I mean, if you're that fatalistic, why worry about AI (or even climate change) in particular? If we're all just doomed no matter what, you may as well just enjoy what you can from life and not stress too much about any particular development.
You're still missing it, we are not in a deadlock. Developed countries are in fact decarbonizing. China too is decarbonizing, though they're behind where the West is, but their goal is to peak their emissions by 2030.
In fact, it's kind of the opposite of what you say—everyone is contributing fractionally to the solution. This is what climate doomers miss.
We are much greener though, at least in the West. Climate emissions peaked in Europe and North America in the last few decades (earlier in Europe.) In Europe, forests are growing back, because marginal agricultural land is being returned to forests as yields rise on prime land. I think this is beginning to happen in the US as well.
This doesn't mean climate change isn't a problem, because even with this progress, we're way behind and not moving nearly fast enough. But often it's the green side of the spectrum that's lying by catastrophizing and understating progress, while overstating the severity of what's happening.
It's happening similarly with AI, where the green movement has decided that AI is unacceptable, even though it has a tiny ecological footprint compared to activities like watching Netflix or eating nuts, let alone eating beef or flying on a plane.
Is the issue that the suggestions from the AI tool are not good, or is that bad code is making it into the repo?
The latter problem should be prevent by code review (first by the developer using the AI tool and then their teammates on a PR.) Code generated by AI should be reviewed no differently than code written by a human. If you wouldn't approve the PR if a person wrote this code, why would you approve it because an LLM wrote it? If your PR process is not catching these issues, you have a PR process problem, not an AI problem.
I think your definition is completely reasonable, but it's not how people answer the question (again, based on reading the report, which has a lot more detail.)
The living paycheck to paycheck stat is meaningless. It generates headlines, because bad news wins, but if you dig in, even the surveys that ask about this reveal that a large percentage of people living paycheck to paycheck are very financially secure, both objectively, and according to their own assessment. I don't have time right now, but I'd encourage anyone reading this to go find the actual source for this stat and read their full report. Last time I looked, it was freely available.
Many of those people living paycheck to paycheck have a budget where they're saving or investing a significant amount of their money, and accordingly, money feels tight. This is the financially sound way to avoid lifestyle creep, but it doesn't mean you're in a precarious position. Or alternatively, they're paying for really expensive, but optional things like private schooling and expensive cars or vacations.
Obviously some people are struggling, but this paycheck-to-paycheck stat is not an accurate way to quantify that. It's best IMO to look at objective metrics like the poverty rate, or people's objective financial picture (available via surveys.)
> Requiring benefit is a huge step from where we are.
The phrase "requiring benefit" is doing a huge amount of heavy lifting in this statement by assuming benefit is something that can easily be agreed upon, or that it's not already taken into account.
Since it will obviously be difficult to agree on this, what if we just vote? I'd propose a system where each of us gets a certain amount of tokens, and then we use our tokens to vote on which products are beneficial. Those companies that make beneficial products will receive the most tokens. They can distribute some of them to their employees and shareholders. The companies that don't provide enough benefit will naturally go under as they fail to accumulate enough tokens to attract employees. We can also set up systems to ensure that everyone receives a minimum number of tokens, and also so that certain people don't accumulate so many tokens as to give them too much control over society.
I believe this system will work well for rewarding companies which produce benefits and punishing those who don't and look forward to it being implemented.