The logical conclusion of this arms race is that eventually they are going to make things so hard that no human will be able to get in without an AI algorithm at which point the only people accessing the site will be the scammers.

...if the allgenations are true the Chinese government might disagree.

Market share is based on the number of active phones in circulation which is effectively the number of activations over several years. The new activations data above is a snapshot of the market share over a just one year. Even if Apple sold zero iPhones last year it would still maintain a reasonable market share because most people do not replace their phone every year.

What about an Early Access game that promises several features you really want and then abondons those promises and just releases as-is? It's easy to see how the new rules can be abused by game devs just like the old rules could clearly be abused by players.

Perhaps with early access games half the purchase cost should be held by Steam and when the game is finally released purchasers can choose whether to get a half-refund and lose access to the game or keep the game and release the held money to the devs. That way it discourages abuse by either devs or users and also provides a lot of encouragement for game devs to actually complete games as promised.

Really? Why? There is no uncertainty given on the measurement and they only quote it to one significant figure which implies that the uncertainty is at least 0.1 making that 0.3 gap less than 3 sigma from zero. Scientifically speaking that's not clear evidence of anything and signficances less than 3 standard deviations disappear all the time due to missing systematic effects or sometimes are even just statistical fluctuations.

It is critical to know the inner reasoning in order to determine whether something understands. A parrot can speak but I do not think anyone believes that it understands what it is saying.

If you understand the concepts behind the words rather than the pattern the words make then you can use logical reasoning to determine new information. An AI trained on word patterns cannot do this and so, faced with a new situation has no clue how to respond and is far more likely to get things wrong. This is why ChatGPT performs so poorly on even simple, first-year university physics questions when asked to explain observations or results...and this is with situations that are known and have happened before. Being able to take concepts and using them to logically extrapolate what will happen in different situations is a key hallmark of intelligence and that is something that current AI simply cannot do.

The point _was_ made: "constant rate" means that the rate of expansion remains the same with time. What you are talking about is a _common_ rate of expansion. The summary says that they are considering variations in the rate as a function of position but, by saying that the rate is constant that inplies that it does not vary with time and that is wrong: we know from multiple supernova studies that the expansion is accelerating. This even gave is a new possible "end of the universe" scenario: the "big rip" where in the incredibly distant future if the expansion keeps accelerating then possibly at somepoint the causally connected region of the universe might shrink to the planck scale at which point space-time itself will become impossible although this is all highly hypothetical since we do not understand what is driving the expansion.

Then we would know because while the universe refuses to define a common zero for velocity (velocity is relative), it absolutely does define a common zero acceleration.

It's already known that the expansion seems to be accelerating. Several supernova studies all point to an accelerating expansion likely driven by Dark Energy but, regardless of the cause, we already know the expansion rate is not a constant.

I suspect that for anything that still needs it you can use an FPGA implementation of the Z80.

Perhaps not but I've a lot of experience using "AI" (although I prefer the term machine learning) for analysis it can and does do stupid things from time to time. In my case that results in some obviously wrong physics results but my code is not in control of a state-of-the-art, fully armed fighter jet.

Exactly how would this apply given that they were protesting _at_ the employer's premises and disrupting other employees who were trying to work there? It seems very reasonable to me that if you turn up at your place of employment and use your access to that place to disrupt the normal business of your employer by staging a sit-in that you should get fired for doing so.

After all, if these people really believed in what they were protesting then the honourable thing to do would be to resign from Google first, like government ministers do when they have a strong moral or ethical objections to the actions of the government of which they are part. Yes, it's a tough decision to make with financial repercussions but if you are not willing to do that then what you have is a preference not a strong moral objection.

I suspect the people who will be most shocked will be those who hold the copyright on the original works that the AI was trained on and is very likely to be partially regurgitating given an output as long as a book.

Yes but Google is not congress and they were not passing a law. That's the problem with the US constitution, unlike moden constitutions that define rights and hence stop anyone who tries to take them away, the US constitution only limits the US government and in today's world large companies often have as much influence on our lives as governments. That being said if you start publicly denouncing your employer it is absolutely reasonable for them to fire you.

The important question is what does the AI in charge of the armed aircraft think it means? One misinterpretation or hallucination and there are going to be a lot of very angry former pet owners.