With a population of 12M and life expectancy of 80 years (give or take), probably not, as that needs ~400 newborns a day to maintain, and the birth rate is probably well below that, like in most Western countries.
> without which, these discoveries do not help much at advancing the field,
At least the fact that frontier models are not optimized primarily on formal math reasoning is in itself is good for not putting mathematicians out of their jobs, isn't it?
How do these results (and the future painted by them) affect your profession; do you expect a similar route as in software, where junior developers are unemployable, LLM-assisted development is the norm, and that great developers stand out partly through better communication with their managers?
The event in this case is a daily observation, so in the last 45 years there have been 16k events. Under Gaussian assumption, 3.5 sigma outliers in positive direction are 1 in ~4,000 events, in either direction 1 in ~2,000, so approx. 8 days would be outliers in the recorded period. Judging by the looks, it's pretty close. And the data is absolutely not Gaussian: at best it's the result of a Gaussian process, as the daily recordings are not independent, but dependent on the previous day.
For completeness, the person who adjusted the thermostat in this case was also hired to get stuff done, and the front door was left unlocked. They still shouldn't have done it, but they were effective.
As the author mentions Siddhartha by Hermann Hesse, let me point out that the (very pretty) edition by Macmillan Collectors' Library is a very poor translation, and the more mundane-looking one from Pelican Classics is an excellent one, paralleling the original German.
> If execution is everything, and frontier LLMs solve execution
Steve Jobs didn't mean "writing good code" as "execution", he meant "making things that align with how people want to use the product". Frontier LLMs are not solving Steve Jobs-level execution (even in the near future) because what that would mean understanding human nature – something that Steve Jobs or Henry Ford were much better than most in the industry today, so at best we're going to see LLMs making things that are as good as the majority of products. Because AI does not see whether something is perfect-according-to-Steve-Jobs or merely good-code-according-to-engineering-practices.
I'd also add a willingness to follow the (perceived) intent of the rules as opposed to gaming the letter of the rule. An experienced folk might say, "yeah it won't matter legally", and continue with "but it will matter to me because this rule is in place for this-and-that reason".
Yes, it and the other three posts sound positively AI written. The first post on the blog is how OP uploaded a backdoored dataset to HuggingFace and left it there for 6 months – whether made up or not, it doesn't sound great.
Why wouldn't one buy a second-hand refurbished MacBook Air M1 with 8 GB RAM and 256 GB storage for the same price? [1]
I really don't see the advantages of getting a newer computer for the same price with worse specs, unless it's a gift and social customs only allow a new one.
Maybe you wanted to visit Mexico and your dream was specifically about Cancún, but then you ended up in Veracruz and were like "oh well, it is Mexico after all, I'd rather be here and visit five other countries similarly than only a single one with my dream city."
I expect that a large portion of the actually – not supposedly – targeted demographic will still not care or know how to set up encrypted comms, and I guess the EP also expects them not to. If someone actually wants to evade CSAR, they probably would know how to (and if not, all the better).
If Chat Control passes, then encryption will not be effortless and usable by the masses, that's the whole point. Basic encrypted chat will be on the level of Snowden trying to communicate with the journalists back in the days – only possible if both parties are willing to go to lengths.
Even assuming good intent, that was extremely silly from them to donate their money away when they don't have anything to begin with. This either assumes that the community is wiser at doing their work (doesn't sound to be the case), or that they were betting that it'll work out one way or another (most likely through another donation) – not realizing that such a donation is exactly what guarantees their organization's future.
I also found woodworking recently as a software engineer and it's incredibly rewarding. Both the tactile feeling of the activity, the idea of building something that _exists_ in physical form and exists in your or a loved one's home, and the pride that you feel about a finished product and having overcome challenges and learned something.
Unlike knitting, I love its usefulness. There are so only many use cases for knitwear, but furniture, man, everyone needs furniture. And being in a home that I built by my two hands is infinite joy.
The three aspects where it falls short to knitting:
- It can't be done mindlessly. It would be unsafe and you'd make costly mistakes that you can't undo by pulling on the yarn.
- It's more expensive. The materials are a bit more pricy (compared to hours spent on working them), but the machines certainly are.
- You are confined to space and time. Whether it's your garage or wood shop where you have machines and can make noise and dust, or it's your living room where you exclusively use hand tools – you surely can't do it in your car while waiting for the kids, or at the university, or on the public transport. Whittling small objects is the one exception.
> “I don’t think this could have happened in any country other than the U.S.,” Dr. Urnov said.
> “We all said to each other, ‘This is the most significant thing we have ever done.’”
And then in the Discover More section is this article:
> Lab Animals Face Being Euthanized as Trump Cuts Research
"Try therapy" is so overly vague that I expect it's nearly useless for a sizeable chunk of people (especially among HN readers).
I've tried therapy with five different therapists in the last seven years. Every time I came away feeling the same, wondering if I'm doing it wrong or if I missed an instruction in primary school.
Can I get by without it? Apparently yes. Am I doing things more optimally with one? Marginally, at best.