I flew Delta about 6 months ago and they had something similar, also for free, but they use Viasat. I think most of the big airlines were moving this way anyway to be honest, Starlink just has a good opportunity for advertising.
I live in what is probably the first place to get these things in the world, but it feels like fiber is being built at an extremely rapid pace. Just in the past couple of years it seems like Google and AT&T fiber went from being a relatively confined thing to being available everywhere in the city, and everywhere outside, and at my friend's ranch 100 miles in the middle of nowhere. Everywhere.
No it should not be, but not because of the dextromethorphan or the phenylephrine being ineffective. By far the biggest issue is the acetaminophen it contains, which it isn't super obvious about, and frequently leads to acetaminophen overdoses. The vast majority of acetaminophen overdoses occur because people combined different medicines containing it (like DayQuil and Tylenol) without realizing they were taking the same thing multiple times. Its a completely preventable cause of liver failure and we should not be making cocktails with it that don't clearly show exactly what they are.
Of course there are paths to success, and they are available, but we live in a tiktok world where everyone wants to be a millionaire tomorrow and not in 20 years.
Its just another symptom of the silicon valley startup "do something illegal, scale fast, and dare the government to do something about it" business model. It is an absolute cancer to society, if there is such a thing as late stage capitalism that is it, and we need to vaporize companies that do it with fines and jail time.
On a broader scale it can be viewed as part of modern culture going towards "screw everyone else, I'm gonna do whatever to get ahead." Kids on tiktok are teaching each other how to scam people. At this rate we'll have scam call centers in the US by the end of the decade and nobody will do anything about it.
It probably is more secure, to be honest. I trust Google to keep my account secure more than I trust some random website to store password hashes and verify securely.
This is really cool, just as I'm starting to get towards the back end of the Kaishi 1.5k deck so this will be perfect for my Japanese studies. Thanks for sharing.
Open-weight models can be abliterated automatically with open source tools though and completely decensored. You can't do that with a closed cloud model.
Not really even in the same ballpark as what they did. These other labs are using AI generated content (which has already been ruled un-copyrightable) to train their models. Oh and they are paying for those tokens. So at absolute worst, they are violating the terms of service. The horror. Meanwhile these frontier AI labs pirated and scraped everything they possibly could, paying not a dime to the copyright owners, nor paying anything to the websites they DDoSed.
Well Zai's GLM 5.2 legitimately is a frontier-level model, though not quite parallel with Opus or Fable. Unfortunately, its too damn big to run locally for most people. Thats the bottleneck right now; the open-weight models exist but something capable of competing with the frontier models just can't run on anything normal yet.
Really cool writeup. I personally like the Cure Dolly approach to conjugation, in which she said there was no such thing as conjugation (at least not in the Indo-European sense), only helper verbs and adjectives. Japanese grammar is just so difficult to understand as a native English speaker. Thanks for sharing.
It almost seems like the juice might not be worth the squeeze. If you want verifiable code that conforms well to a well-designed plan, you have to basically write pseudocode and have the AI translate it for you. At that point why use the AI to write the code at all? And then, personally, I find that I just have more fun planning, writing, and debugging myself. I think its kinda the part of programming that I fell in love with in the first place.
We've also seen ample evidence that AI labs are not overly concerned with the legality of how they obtain training data. Its not a stretch to say maybe they look at some other stuff they shouldn't too.