I would say, whichever one is polling highest, who has signaled support in any way for the issue. Other things about them don't matter because they aren't going to win anyway. In this case a lot of third party votes for a single candidate are probably better than the same number of votes spread across different candidates because it looks more like an organized voting block to politicians looking at the numbers in retrospect, who are the reason to vote at all if you are voting third party, you are trying to communicate via those numbers about how your vote can be obtained or lost.
There's no telling how much it would be worth since there's no active market for it, the NFT has been owned by the same wallet since its purchase in 2021 and has not been transferred or resold. That wallet currently has minimal value in Ethereum or tokens, and around 2k different NFTs, most of which don't seem to be very valuable. They are still active, with a transaction from a month ago moving 264k in stablecoins to a crypto exchange.
I thought it was the electrons that orbited
If you are saying they will privatize elections somehow as a loophole around following the constitution, I am skeptical
She's got the horns, is that her father then?
The right of citizens of the United States to vote shall not be denied or abridged by the United States or by any State on account of sex.
Only about a century old. Let's go for another 100.
If it wasn't clear, I am not claiming that AI is better than a person at summarizing complex information.
The AI summaries were judged significantly weaker across all five metrics used by the evaluators, including coherency/consistency, length, and focus on ASIC references. Across the five documents, the AI summaries scored an average total of seven points (on ASIC's five-category, 15-point scale), compared to 12.2 points for the human summaries.
The focus on the (now-outdated) Llama2-70B also means that "the results do not necessarily reflect how other models may perform" the authors warn.
to assess the capability of Generative AI (Gen AI) to summarise a sample of public submissions made to an external Parliamentary Joint Committee inquiry, looking into audit and consultancy firms
In the final assessment ASIC assessors generally agreed that AI outputs could potentially create more work if used (in current state), due to the need to fact check outputs, or because the original source material actually presented information better. The assessments showed that one of the most significant issues with the model was its limited ability to pick-up the nuance or context required to analyse submissions.
The duration of the PoC was relatively short and allowed limited time for optimisation of the LLM.
So basically this study concludes that Llama2-70B with basic prompting is not as good as humans at summarizing documents submitted to the Australian government by businesses, and its summaries are not good enough to be useful for that purpose. But there are some pretty significant caveats here, most notably the relative weakness of the model they used (I like Llama2-70B because I can run it locally on my computer but it's definitely a lot dumber than ChatGPT), and how summarization of government/business documents is likely a harder and less forgiving task than some other things you might want a generated summary of.
It can also be a solid rubber duck for debugging.
A lot of the time I get 3/4 of the way through writing a prompt and don't bother hitting enter because I already figured it out. Great way to get your thoughts organized to have an incentive to put them down in writing.
If other people are also immortal, the awkwardness of all of them eventually becoming your exes
IMO the most valid argument is that there are way more people making a middling income than people making a high income, so any reduction in taxes for those people would need a proportionally much larger increase in the upper brackets to maintain the same level of tax revenue, if it's possible to make the numbers work at all depending on how much of a tax break you want to give. The minimum amount to be taxed is set based on where the tail end of the bell curve is, the number of people who are poor enough not to be taxed is small.
Of course there's also the fact that the richest people don't get their money from having a job at all, it's all in investments, so messing with income tax rates doesn't even affect them.
But I think the point is, the OP meme is wrong to try painting this as some kind of society-wide psychological pathology, when it's rather business people coming up with simple reliable formulas to make money. The space of possible products people could want is large, and this choice isn't only about what people want, but what will get attention. People will readily pay attention to and discuss with others something they already have a connection to in a way they wouldn't with some new thing, even if they would rather have something new.
“What I said is this, let’s be frank, I said encourage the criminal to fight, encourage them to draw their guns. That was my instruction, encourage them to fight, and if they fight, then kill them so my problem in my city is done,” he said
Earlier this month, a separate parliamentary inquiry heard evidence from a former police colonel who said officers could earn between 20,000 pesos (£265) and 1m pesos (£13,200) per killing during the drugs crackdowns, depending upon the target. Rewards were given only for killings, not arrests, she said
Duterte denied authorising police to kill suspects, saying he had never ordered his national police chiefs to carry out extrajudicial killings.
What a ridiculously flimsy way to deny responsibility
Nice, how do we make it happen?
The biggest reason that is often overlooked is wealth inequality. The rich keep accumulating wealth, and real estate is a scarce form of wealth that holds value, produces a return, and can be accumulated. It probably accelerated recently because of the large amount of money that was dumped into the system around covid; that was yet another opportunity for the wealthy to grab a bigger share of the pie.
If things keep going this way, we're going to get into a situation where regular people don't own houses anymore, and rent is a much larger percentage of your income.
IIRC the story this is from is about a girl who is incapable of making the most basic decisions so an assistant was hired to whisper in her ear what she should be doing
I wrote off politics media as hyperbolic and manipulative propaganda in 2016 and I actively distance myself from it, so I've only seen the broad strokes of this current election cycle. Unless you honestly believe you are doing important activism work, give yourself permission to just chill out about politics. If your life is full of problems caused by politics such that it's impossible for you to chill out about politics, you have my sympathy.
that is not the ... available outcome.
It demonstrably is already though. Paste a document in, then ask questions about its contents; the answer will typically take what's written there into account. Ask about something you know is in a Wikipedia article that would have been part of its training data, same deal. If you think it can't do this sort of thing, you can just try it yourself.
Obviously it can handle simple sums, this is an illustrative example
I am well aware that LLMs can struggle especially with reasoning tasks, and have a bad habit of making up answers in some situations. That's not the same as being unable to correlate and recall information, which is the relevant task here. Search engines also use machine learning technology and have been able to do that to some extent for years. But with a search engine, even if it's smart enough to figure out what you wanted and give you the correct link, that's useless if the content behind the link is only available to institutions that pay thousands a year for the privilege.
Think about these three things in terms of what information they contain and their capacity to convey it:
-
A search engine
-
Dataset of pirated contents from behind academic paywalls
-
A LLM model file that has been trained on said pirated data
The latter two each have their pros and cons and would likely work better in combination with each other, but they both have an advantage over the search engine: they can tell you about the locked up data, and they can be used to combine the locked up data in novel ways.
Tool to copy Reddit comment chains
So I was reading this post and decided to make the tool described, as a userscript (I credit ChatGPT with doing most of the work, which went pretty quickly). To use it, install a compatible userscript browser extension such as https://violentmonkey.github.io/ , then press install on the linked page. Reddit comments should now have a 'copy-context' button that will put the comment chain in your clipboard. I made it for old.reddit so probably won't work with the redesign. Another limitation is that it will only work to copy what is on the current page, so if the comment chain is too deep it's not going to get all of it.
Any feedback is welcome. Also if someone who can read javascript wants to give it a once-over and confirm for people that it isn't malicious that would be cool too.