> Microsoft’s security researchers in the fall observed individuals they believe may be linked to DeepSeek exfiltrating a large amount of data using the OpenAI application programming interface, or API, said the people, who asked not to be identified because the matter is confidential.
> Such activity could violate OpenAI’s terms of service or could indicate the group acted to remove OpenAI’s restrictions on how much data they could obtain
What do we think this means in practice?
"Exfiltrating data" makes it sound like they were taking private chat logs, but I imagine that would be a much bigger deal. I'm assuming it's just using multiple free OpenAI accounts across a bunch of different IP addresses to generate a large training set.
The article is (I'm assuming knowingly) written to give the impression that this was the work of some elite state-sponsored hackers exploiting vulnerabilities in Microsoft/OpenAI's software. In reality they entered their credit card info and typed some commands, same as everyone else.
Just big sounding words to make it sound like something nefarious is happening. Nothing of that sort actually happened, just OpenAI trying to save face.
Right, Congress was shown some pretty convincing evidence that execs in China pull the strings, and those execs are vulnerable to Chinese government interference.
As we’ve seen in the past couple of weeks, social media companies based in the US are also vulnerable to US government interference — but that’s the way they like it.
They released a Marty Rimm-level report citing that pro-Palestinian was mentioned more than pro-Israeli content in ratios that differed from Meta products. This was the 'smoking gun' of manipulation when it's more of a sign Meta was the one doing the manipulation.
The opinion today has almost nothing to do with how content is controlled on the platform; the court is very clear that they'd have upheld the statute based purely on the data collection issue.
I don't know what Congress has said but there absolutely is evidence that TikTok has been used to spy on users for political reasons. A US based engineer claims that he saw evidence that Hong Kong protestors were spied on in 2018 at the behest of a special committee representing the CCP's interests within ByteDance. This is not surprising, most major corporations within China maintain a special committee representing the government's interests to company executives
Do law enforcement portals provide current location information? There's an extended history of the TikTok being used to spy on the location of user devices
Okay, that's because Grindr users choose to publicly share their current location; that's the point of the app. Governments having an API that lets them access data that users publicly share seems substantively different from governments having access to private information, obtaining that information by subverting internal controls at TikTok and ByteDance intended to keep it private. I think anyone not arguing for arguments sake would acknowledge that
Most apps coerce their users into sharing location information. That's why they released apps and did not just use progressive web apps in the first place.
But, this is done under the guise of commercial interests, usually advertising, so it's okay?
That's the way I like it for my children. Pardon the demagogue. The US, being the awful mess it is is still 100x better IMHO than the chinese government. It's the lesser evil kind of thing and honestly the reason I believe that democracy is 100% THE way to go. Things can only get US level nefarious with democracy. Far from perfect but much less evil.
The only problem with democracy is that it's so fragile and susceptible to bad non-democrat actors intervention, which is more of an awareness problem.
No, X doesn't have a corporate governance structure that requires Chinese government control, because it is a US company.
Companies in China (and especially those of prominence) have formal structures and regulations that require them to cooperate with the government, and sometimes require the companies to allow the government to intervene in operations if necessary.
It is not possible for a CCP official to show up to a board meeting at X and direct the company to take some action, because that isn't how US corporations work.
A CCP official could show up at a Tesla board meeting and announce they're going to seize Gigafactory Shanghai unless Musk takes down some content on X. There doesn't seem to be much of a difference.
Tesla is quite notable as the only foreign automaker which China has allowed to operate independently in China. All of the rest of them were forced to joint venture with 51%+ control being handed over to a Chinese domestic company. So, really it's pretty surprising that they haven't done that even before Musk owned X.
But regardless, there is a huge difference between a request and actually having managerial authority -- the most obvious being that someone with managerial authority can simply do whatever they want without trying to compel someone else. Also, X, being subject to US law, must comply with that no matter what consequences Musk is threatened with. So, any threats may have limits in what they can practically accomplish.
You are assuming a lot about supposed evidence nobody has said anything specific about. One shouldn't also assume people in Congress know how to evaluate any evidence. Nor justices, based on the questions they asked.
As a matter of political science and public choice theory, the legislature is the branch of government most trusted to collect information and make these kinds of deliberations.
Would you call Marjorie Taylor Greene a qualified and trusted investigator for the american people? I sure wouldn’t. Talking about what the legislature is supposed to be is irrelevant. What the legislature actually is is relevant.
You might buy that, but I don't. Unless they can actually put forward publicly compelling evidence of a national security risk, this can only be seen as a handout to Facebook by the government. This saga just gives more evidence that the US government exists primarily to serve the interests of US's oligarch class. Aside for those oligarchs, it does nothing to serve US citizens' interests.
Congress members speak of space lasers and weather control... I'm not sure they're competent as a whole. Actually, it reminds me of the Russian guy that always spouts nonsense about nuking UK into oblivion, and that theory that he's just kept around to make the real evil people look sane.
I’m sure he’s bending at the knee right now because he feels very secure and just had a change of heart about everything precisely one month after the election.
Is he bending the knee, or dropping the mask? The billionaire+ class rightly sees this as their big opportunity to seize power for the next several generations, removing worker and consumer protections and enshrining themselves as essential parts of the government.
> Of course, it only took a couple more years to get photorealistic image outputs.
"Photorealistic" is a pretty subjective judgement, whereas "does this code produce the correct outputs" is an objective judgement. A blurry background character with three arms might not impact one's view of a "photorealistic" image, but a minor utility function returning the wrong thing will break a whole program.
I know you're joking but technically it does have AI, the SOC is built on Nvidia's Ampere architecture with tensor cores. If nothing else they'll probably be used for DLSS upscaling.
The even harder truth not mentioned here is that existing tools have a hard time understanding large codebases with well-establishd internal patterns and libraries.
The article mostly talks about how AI tools can help with new things, but a large amount of software development is brownfield, not greenfield.
This is not a problem at all as long as you use very good typing because the local contract boundaries are what matter unless you use huge amounts of global state which everybody knows is a very very bad idea and has been demonized for decades
Totally. If I would tell 10-year-old me that I have my own website on my own domain, it would be seen as a shocking development. I find it really cool to be able to have a corner of the internet that is just mine.
You don't even need attention - it could be a public blog where you share about the things you learn. You never know who it's going to help. That's primarily my motivation with the blog.
Sometimes projects get stars just because people like the personality or company behind the project.
Case in point: https://github.com/facebook/hhvm/. It got 15,000 stars in its first few years, but roughly 10 non-Facebook companies actually ever used it in production, and today only one non-Facebook company uses it (I work at that company).
That doesn't mean that the stars are just because people like the company. People may find the technology interesting even if they have no intent of using it.
> Such activity could violate OpenAI’s terms of service or could indicate the group acted to remove OpenAI’s restrictions on how much data they could obtain
What do we think this means in practice?
"Exfiltrating data" makes it sound like they were taking private chat logs, but I imagine that would be a much bigger deal. I'm assuming it's just using multiple free OpenAI accounts across a bunch of different IP addresses to generate a large training set.
reply