Ben Thompson has been covering Intel’s precarious position for over a decade (well before the market finally realized it) and the latest update is not looking good:
Intel’s is technically on pace to achieve the five nodes in four years Gelsinger promised (in truth two of those nodes were iterations), but they haven’t truly scaled any of them; the first attempt to do so, with Intel 3, destroyed their margins. This isn’t a surprise: the reason why it is hard to skip steps is not just because technology advances, but because you have to actually learn on the line how to implement new technology at scale, with sustainable yield. Go back to Intel’s 10nm failure: the company could technically make a 10nm chip, they just couldn’t do so economically; there are now open questions about Intel 3, much less next year’s promised 18A.
I know that post, but the problem is he is just extrapolating from history. Not a bad thing in absence of real information, but... Well, let's hope he's wrong. :-)
In short, the Go module proxy causes an excessive traffic volume on git VCS sources with frequent clones of unchanged repos. Regardless of whether or not the developer is/was always reasonable in how he discussed this, he was absolutely right about this being a hostile behavior from the official Go proxy that is the result of bad/insufficient engineering. The team's suggestions to simply stop refreshing his one domain were also not sufficient given that the problem clearly impacts all Go module VCS hosts.
The developer also appeared to be banned in a way that violated the Go CoC's own provisions around fair notice and a proper hearing, which is super disappointing to see.
Oh man, was Drew banned from all Go spaces, or just from the issue tracker as he mentioned? He seems to draw ire, although whenever I actually read what he writes, he usually makes a lot of sense. I imagine there are examples of him being abrasive, but it usually seems like he values being thoughtful and kind.
River ( https://riverqueue.com ) is a Postgres background job engine written in Go, which also has insert only clients in other languages. Currently we have these for Ruby and Python:
Having recently adopted Resend and skimmed a bunch of different email APIs, I'm still waiting to find a single provider whose API supports Stripe-style idempotency [1] so that I can guarantee I don't send the same email through their API multiple times. I'd like to confidently avoid accidentally spamming a user if i.e. a background job retries multiple times due to an unrelated error, or merely from failing to receive the API response that an email was sent/created successfully.
I don't even expect them to keep a list of IDs forever, just some best effort like "we don't send anything with the same id twice in a one-hour window"
Gmail's API did support this I believe, and then ofc my org transitioned to Microsoft so my little homemade email service quit working
While you are here, I have been asking for years to have better access control on API keys so they can only use assigned servers. So my staging cant send prod emails...
I understand this change might be a large undertaking to implement across the entire API, and although it would be good to do everywhere, it’s primarily just the send email API that needs it.
Thank you! I'm not sure how this didn't come up in what I felt was a pretty comprehensive round of Google searching, but it looks like a fantastic option. If only they had free allotment or something <$20/mo for a small number of emails :) But otherwise this looks a the solution I was looking for.
This is exactly what I've been wrestling with recently.. it's so disappointingly absent in all the offerings out there.
The solution I ended up with was to build my own pseudo-idempotency around Postmark. It helps that Postmark at least has proper persistence so you can query their API to check if you've already sent a certain email. I had to move away from Mailchimp because, if an email gets queued for some reason, it isn't reflected as sent in the API so there's no way of knowing whether an email is hiding in "queued limbo". Postmark doesn't have this problem.
The only caveat is that there is a delay between sending an email and it being reflected in the API (seems pretty standard across the different services). This means I had to implement a simple database backed lock to make sure I never try to send the same email more than once in quick succession.
It's kind of ridiculous, but.. I couldn't find any alternative.
Edit:
Wonder if it would be worth wrapping something like this up into some kind of a self-hostable proxy service. You'd provide the idempotency key in a header and it would do the rest
Resend is also built on top of SES, and yet it's an extra layer of hosted services on top of it. So this is actually a self-hosted alternative to i.e. the functionality that Resend offers, even though there are still other underlying dependencies. Seems like a reasonable way to describe it IMO.
It seems like the author already went above and beyond to try to get this resolved through the proper channels. Why did it take an HN front page post to get it actually looked into? It does not inspire confidence in your service.
The thing is that Cloudflare has no proper channels. Anyone that has tried Cloudfare's support (even as a paying customer) knows that it's almost impossible to get a sensible answer to anything.
That's the way tech companies work today. You need to have connections or spend social capital to reach people who can solve issues. Or you need to bang your head in support channels for weeks hoping that someone will escalate your issue.
Sometimes there's workaround, if you pay big bucks, you might get "personal" manager who can actually connect you with necessary people. But that service is not available for every company out there. And if you don't pay big bucks, you don't have a chance.
This scheme probably makes sense. At certain scale you just can't talk to everyone. When you have 20 developers and 20 million clients, one person can have only so much time. And most support issues are stupid anyway.
I recall several similar situations with Cloudflare: a user has billing/service problems and the only way to get some support is to end on HN frontpage or reddit ...
I remember that post and I’ve read it a few times, thank you for it! I was already working on River at the time but it was refreshing to see the case made so strongly by another person who gets it.
> But a small subset of mostly right-wing reactionaries
If you look at the photo of the protestors in the article or research the groups stated to be involved, it seems pretty unlikely that these individuals are "right-wing". There is a contingent in support of this that is likely more conservative by way of being law enforcement aligned, but there are also other segments in support of these measures that have very different motivations.
I am not one of these people FWIW—I think opaque on-device scanning is a clear and unacceptable invasion of personal privacy.
I don't accept your "just look at them" photo analysis to be any real indicator of the political leanings of the movement, especially not when these people appear to be the few opportunists in the Cupertino area who physically showed up. I will instead rely on tangible data that that suggests that these kinds of causes are propelled largely due to right-wing reactionary politics and the moral panics they cultivate.
In the United States at least, the MAGA right tends to be rather obsessed with pedophilia and child sex abuse, and use it as a cudgel to attack any politician, business, or group of people they don't like.
They call Joe Biden a pedophile, and use photoshopped images to suggest that he has a penchant for touching children. It's not true, and his dishonest accusers are the ones photoshopping children into implied sexual situations, but the lie persists because it is useful
They call Transgender people pedophiles and "groomers", and suggest that they should be forced from public life and treated like criminals, being denied any job that might put them in contact with a child. There is no evidence that queer people are any more likely than the general public to engage in child sex abuse, but the lie persists because it is useful.
They even managed to stir themselves into a frenzy with the claim that a pizza restaurant in D.C. that was popular with members of the Obama administration was actually a front for a child sex trafficking ring masterminded by Hillary Clinton. There's absolutely zero evidence for any of that, but the lie persists because it is useful.
While there are definitely people that fall for this propaganda that aren't right-wing, right-wingers make the perfect useful idiot for this kind of cause.
The people at the top of the chain. The people who made those professional looking signs. The people who run the OSEAC. They are predominately right-wing, and have political goals that are far divorced from any notion of "protecting children".
As a general rule of thumb, in America, if there is any group of people trying to convince you to give up your rights and freedoms because children are in danger, if you follow the breadcrumbs, you'll find Republicans in charge.
On a side note, for an or that wants to fight back against the exploitation of children, the Heat Initiative having a website covered with AI-generated images of children is horrible optics, given how Generative AI has become a focus on CSAM creation.
If you impersonate another actor for profit claiming to be him/her you may be facing legal action. There is a difference in selling yourself as "Scarlett Johansson" vs. "A Scarlett Johansson impersonator" vs. your own name/stage name. If it's covert and deceiving, you are in trouble, if it is overt, you are fine and that's why they world has so many Elvis impersonators. Although I see a new market opening for Kate Moss impersonators https://news.ycombinator.com/item?id=40498668
When it comes to works of art, if you tell the buyers/viewers it's yours you are fine, but if you paint a copy of a famous painting and claim it to be the original one, you are a criminal.
I'm not a lawyer, so I don't know if this is relevant or not, but it's not really Scarlett Johansson's voice we're talking about, it's "the character in Her that Scarlett Johansson voiced". I'd be surprised if Scarlett's natural voice matches the one in Her - I assume she would have changed her pitch and/or cadence for the movie to match the director's requirements. Samantha's voice in Her has a subtle, 'not-quite-natural' sound to it, which would make sense for a computer-generated voice.
Can the movie studio that produced Her claim copyright over Samantha's voice? If they release "Her 2" without Scarlett Johansson, could the movie studio hire someone who sounds like Scarlett to play Samantha's character?
I don't think this case is as simple as people make it out to be. Even Sam Altman's tweet: "Her", seems to confirm they took inspiration from the movie Her, so is it wrong to hire a voice actor that sounds like the character in the movie? If Sam had tweeted, "Scarlett", then it would be very different.
You were asking about a hypothetical situation in which you may find yourself. That's what I addressed in my reply. The person whose voice was used does not pretend to be SJ afaik. It's the communications between OpenAI and SJ and Altman's tweets that created an ambiguous situation similar to someone trying to ask an attractive person out, being told it's not going to happen, but publicly winking and suggesting that there is something going on. It backfired. OpenAI and Altman are trying to get out of it, but their time of being the darlings of the media is over. There will be more scrutiny of their actions.
Actors use stage names all the time because the actors union doesn't allow duplicate names. Emma Stone's birthname for example is Emily Stone, who is another actor. Actors and musicians are also known to seek as much uniqueness as possible to be able to trade on that uniqueness. If you were a talented singer or painter you'd have trained early on to develop your own distinct voice or style, it's how the music and art schools work.
Fairly sure I understand the public details about this situation. We have at least one newspaper that has reviewed recordings or footage from the voice actor and confirmed that the OpenAI voice is a direct match for her actual regular speaking voice. We also have records illustrating that the CEO had no direct involvement in the hiring of the voice talent, and that there was no direction given for her to "sound like 'Her'".
Without any evidence that OpenAI set out to intentionally find someone who sounded like a fictional character, why should this individual be prevented from seeking work as a voice actor in any capacity?
They wanted Scarlett Johansson specifically, couldn't get her and then hired someone
They hired the Sky voice actress well before contacting SJ.
with the intent to convince the public it was her.
The only evidence for this is Sam's "her" tweet, which can easily be referring to the concept of a realistically-voiced AI assistant rather than SJ in particular.
> They hired the Sky voice actress well before contacting SJ.
Then why would they need to contact SJ 2 days before the release? If they already had the Sky voice complete? If there wouldn't be enough time to have SJ record everything for the voice?
To me, it doesn't pass the sniff test. Either their talent department is just really bad at planning and had no clue it was releasing just two days later (unlikely), or they were trying to get an agreement signed with her before the release, anticipating that this would happen (likely).
Then why would they need to contact SJ 2 days before the release?
According to OpenAI, they wanted SJ for a separate voice (https://openai.com/index/how-the-voices-for-chatgpt-were-cho...), possibly as a "one more thing" conclusion to the demo. You can say that they're lying, and certainly OpenAI doesn't have the highest credibility at the moment, but it's at least plausible.
or they were trying to get an agreement signed with her before the release, anticipating that this would happen (likely)
If they anticipated this would happen, why wouldn't they just ditch the Sky voice altogether?
> If they anticipated this would happen, why wouldn't they just ditch the Sky voice altogether?
This is my opinion so it's not factual. They already don't have any qualms making billions of dollars operating with a maybe-illegal-but-definitely-not-cool business model that hoovers up copyrighted content without compensated authors to use their works to train models to reproduce similar works, again without compensating any copyright owners. Why would they act different when it's one celebrity? Sugar Daddy Microsoft has enough money and lawyers to throw at SJ to keep her busy until she either croaks or goes broke.
Also, why would MS or OpenAI lose out on a billion dollars in stock growth with the release of ChatGPT4.o when, for a few million dollars, you can do whatever you want.
I don't actually see an ethical issue with this situation, unless the voice was specifically trained using clips of Scarlett Johansson's actual voice.
I should have some say over how my actual voice is used. If I am a public figure, ala Donald Trump, and there are a lot of people out there who do a great impression of me, those voices belong to the people actually creating them, not me.
If it turns out that there are people out there who either naturally sound like Scarlett Johansson, or who can alter their voice so that they sound like her, they should have the freedom to profit from their own talents (natural or otherwise).
If the resulting AI voice isn't actually claiming to be Johansson and isn't trained using her actual voice, I don't see anything wrong with it.
"
The appellate court ruled that the voice of someone famous as a singer is distinctive to their person and image and therefore, as a part of their identity, it is unlawful to imitate their voice without express consent and approval."
Thanks for sharing. I completely disagree with their rationale and decision but it’s good to know about the precedent here. I’d be completely on Ford’s side here too!
I would not be surprised to see that precedent overturned at some point because its outcome seems to deny individuals the right to profit from their own abilities, just because they happen to sound like a famous person.
In a novel case of voice theft, a Los Angeles federal court jury Tuesday awarded gravel-throated recording artist Tom Waits $2.475 million in damages from Frito-Lay Inc. and its advertising agency.
The U.S. District Court jury found that the corn chip giant unlawfully appropriated Waits’ distinctive voice, tarring his reputation by employing an impersonator to record a radio ad for a new brand of spicy Doritos corn chips.
The biggest problem is that they were making insinuations that it was intentionally similar to her role in the movie Her (and even contacted her directly to do it).
There are laws for unauthorized use of likeness that could possibly apply here.
I'm not familiar with the specific laws that are relevant here, but I have trouble buying the argument that this constitutes a "likeness" when Johansson was in no way involved in its creation, other than by providing inspiration.
So they asked her and she turned them down. But guess what, they can get a pretty good result without even involving her. It turns out that her "talent" for the voice from the movie Her is in fact not so unique and valuable because it can be reasonably well replicated by other people, so why shouldn't a company be able to choose an alternative like that? A person should not have an exclusive claim to the rights of all voices that sound remotely like theirs.
Celebrities have some right to publicity, so in this case she could argue that they were watering down the value of her as an actress by widely selling an unauthorized sound-alike.
The laws around this aren't strong (which is why her response pointed out the need for legislation), and as I mentioned OpenAI would have been in a much stronger position if they didn't directly reference the film publicly several times. One could argue that they were using the film and her role to promote their product.
ethics and legality are completely different things, this case isn't black and white at all.
if you're just talking ethically though, i still find it shady. it would be like asking Gaudi to design a new building, then when he says no, finding Bob who can replicate Gaudi's style really well. there, it's legal, and you could say it's ethically fine because Bob should have freedom to profit from his skill. But it feels gross
Intel’s is technically on pace to achieve the five nodes in four years Gelsinger promised (in truth two of those nodes were iterations), but they haven’t truly scaled any of them; the first attempt to do so, with Intel 3, destroyed their margins. This isn’t a surprise: the reason why it is hard to skip steps is not just because technology advances, but because you have to actually learn on the line how to implement new technology at scale, with sustainable yield. Go back to Intel’s 10nm failure: the company could technically make a 10nm chip, they just couldn’t do so economically; there are now open questions about Intel 3, much less next year’s promised 18A.
https://stratechery.com/2024/intel-honesty/