maeil

18 ms·

181.
by maeil 2y ago
It doesn't; it makes the ban more likely if anything. See how on certain topics the censorship immediately increased as Musk took over Twitter.
182.
by maeil 2y ago
> US President Donald Trump has also announced that the USA will not introduce the global minimum tax. Corporations are the legal structure Well well well, who'd have thought! 3 months ago on "Italy stiffens terms of digital se
183.
by maeil 2y ago
IME the Swiss national football team has more Albanian and Serbian last names than Italian, and that's usually a pretty good proxy. They clearly managed to get residency. Easier for the kids of the foreign parents?
184.
by maeil 2y ago
Even when preceded by "champagne", the word "socialist" shouldn't be allowed to appear in the same sentence as her name.
185.
by maeil 2y ago
I'm stunned they put Korea on the list, given this is dressed up as an anti-China measure, and Korean leaderships flipflops between anti-China and pro-China every ~4 years when the opposition replaces the government. There's also
186.
by maeil 2y ago
They did not - Switzerland valiantly guards plenty of dodgy assets for the US oligarchy all the same.
187.
by maeil 2y ago
> All companies are greedy. That's the reason for existence. Black and white oversimplification that muddies the water. You really can't think of two companies with incredibly different levels of "greed"?
188.
by maeil 2y ago
Agreed that massive fines (or, in my view preferably, massive stock dilutions) are better. However > You're proposing sending half of America to jail if any Fortune 500 commits a crime. That's a get-out-of-jail-free card, not m
189.
by maeil 2y ago
I own stocks in several companies, as most people on here probably do. There's nothing wrong about what they're saying, I'm a shareholder. If one of the companies I've invested in commits genocide I don't think it w
190.
by maeil 2y ago
The current solution has been leading to the worst possible end .
191.
by maeil 2y ago
> It is perhaps not overly hyperbolic to state that the challenge of our era is to determine a mode of capturing the power of emergent technology to undo this situation, and nothing less than the fate of the planet rests on doing so. Eug
192.
by maeil 2y ago
No, that's not why "we" don't like to do that, as there's countless way around that problem. They don't like to do that because it hurts the main stakeholders: those holding the reigns, the 0.1%. There's
193.
by maeil 2y ago
It has, since day one. The fact that the WIH was one of only 6 institutions in the world that had been conducting GoF research on bat-related pathogens before the pandemic gives such strong statistical evidence towards it that any theory ho
194.
by maeil 2y ago
Your memory is accurate. Yet elsewhere we have an infectious disease researcher claiming > At the same time, I've come to distrust many of the voices who push the lab leak hypothesis, either because they're obviously doing so f
195.
by maeil 2y ago
Similarly: In 2004, nine people were infected with Sars and one person died after two researchers were separately exposed to the virus while working at the Chinese Institute of Virology in Beijing. In November 2019, just a month before the
196.
by maeil 2y ago
How many times do we need to repeat "lab leak != intentionally lab engineered"? Your most relevant parts are all about the latter. You know this, yet muddy the waters.
197.
by maeil 2y ago
> or because they've become addicted to being "the lone voice in the wilderness", despite it not being a risky position to take. This frankly makes me distrust you; in 2020-2022 this was absolutely a risky position to tak
198.
by maeil 2y ago
I'd love to do such a quiz -I might even be willing to pay for the privilege! I'm quite convinced I'm really accurate at calorie estimation without using a scale but would love to be proven wrong. Zero food industry experienc
199.
by maeil 2y ago
Not my point, never said it is. Substitute that example with another criminal act. Edit: Changed it just for you
200.
by maeil 2y ago
> and we really deserve everything we get for it. Those people who conflate them deserve it. You and me don't. > there's really not much you, I or anyone else can do We can make our own community. And Bluesky is very much no
201.
by maeil 2y ago
Didn't know Meta paid for them or borrowed them from libraries. In fact, I don't think they did.
202.
by maeil 2y ago
Once in >5 years is nothing like a SaaS, especially since it's such a tiny part of a product. Even with the traditional software there'd be no guarantee it would keep working for longer than that, e.g. on new OS versions, espec
203.
by maeil 2y ago
> In the most recent fiscal year, Alphabet's net income amounted to 73.7 billion U.S. dollars Absolutely no way. Yup. > Buying millions of ebooks online would take a lot of effort, downloading data from publishers isn't a th
204.
by maeil 2y ago
Correct, but the tail of VC being fat is unique to the US. Pretty much the rest of the world is like the UK in all but a miniscule percentage preferring stability rather than moonshots.
205.
by maeil 2y ago
You didn't read what I quoted and replied to. Please do so before responding with "false".
206.
by maeil 2y ago
This isn't news, the other popular benchmarks are just as gamed and worthless, it would be shocking if this one wasn't . The other frontier model providers game them just as hard, it's not an OpenAI thing. Any benchmark that
207.
by maeil 2y ago
> Do people actually think OpenAI is gaming benchmarks? Yes, they 100% do. So do their main competitors. All of them do.
208.
by maeil 2y ago
"Their model" here is referring to 4o as o1 is unviable for many production usecases due to latency.
209.
by maeil 2y ago
This just isn't accurate, on the overwhelming majority of real-world tasks (>90%) 3.5 Sonnet beats 4o. FWIW I've spoken with a friend who's at OpenAI and they fully agree in private.
210.
by maeil 2y ago
Exact same story for Korea. Dollar-term salaries similar to the EU, but when you compare to CoL it's a much better deal.

More ›