Jasondells

translation:

AI: Privacy Guarantor asks DeepSeek for information Possible risk to the data of millions of people in Italy

The Guarantor for the protection of personal data has sent a request for information to Hangzhou DeepSeek Artificial Intelligence and Beijing DeepSeek Artificial Intelligence, the companies that provide the DeepSeek chatbot service, both on the web platform and on the App.

The Authority, considering the potential high risk for the data of millions of people in Italy, has asked the two companies and their affiliates to confirm which personal data are collected, from which sources, for which purposes, what is the legal basis of the processing, and whether they are stored on servers located in China.

The Guarantor also asked companies what kind of information is used to train the artificial intelligence system and, in the event that personal data is collected through web scraping activities, to clarify how users registered and those not registered to the service have been or are informed about the processing of their data.

Within 20 days, companies must provide the Authority with the requested information.

Rome, January 28, 2025

https://asiatimes.com/2024/10/huawei-uses-tsmc-loophole-to-b... https://m.koreaherald.com/article/3399758

There have been talks about a laissez-faire attitude regarding cybersecurity at OpenAI for a long time.... but this is surely coming to an end now. Same at Google.

The claim that R1 was trained for under $6M on 2,048 H800 GPUs always seemed suspicious. Efficient training techniques can cut costs, sure—but when OpenAI, Google, and Meta are all burning hundreds of millions to reach similar benchmarks, it’s hard to accept that DeepSeek did it for pennies on the dollar. Then Alexandr Wang casually drops that they actually have 50,000 H100 GPUs… what happened to that “low-cost” narrative? If this is true, it's not efficiency—it’s just access to massive hidden compute.

The stolen OpenAI data theory is another red flag. OpenAI researchers have been hit by multiple security breaches in the last few years, and now we have a former OpenAI engineer found dead under very weird circumstances. Coincidence? Maybe. But corporate espionage in AI isn’t some sci-fi plot—it’s very real, and China has been caught running large-scale operations before (Google exfiltration cases, the ASML trade secret theft, etc.).

And then there’s the CCP-backed propaganda angle. This part is almost too predictable—China hypes up a “homegrown” breakthrough, gets state media to push it as “proof” they’ve surpassed the West, then quietly blocks foreign scrutiny. Lei pointed out that DeepSeek won’t even let U.S. phone numbers register. Why? If R1 is truly open-source and transparent, why limit access? We’ve seen this before with ByteDance, Alibaba, etc.—government-approved success stories that follow a controlled narrative.

But despite all that skepticism… R1 is real, and the performance numbers do exist. Whether they’re running stolen training data or smuggled GPUs, they’ve built something that competes with OpenAI’s o1. That’s still impressive. The question is how much of this is a real technological leap vs. how much is state-backed positioning and/or cutting corners.

So what happens next?

If DeepSeek is serious, they need outside audits—actual transparency, full datasets, external verification. Not just another “trust us” moment. The U.S. needs better export control enforcement… we’re seeing massive loopholes if China can stockpile 50K H100s despite all the restrictions. AI labs (OpenAI, Anthropic, etc.) need better security. If OpenAI’s data really did leak, this won’t be the last time. I don’t think R1 itself is a scam, but the surrounding story feels curated, opaque, and suspiciously convenient. Maybe DeepSeek has built something remarkable, but until they open the books, I can’t take their claims at face value.

First, OpenAI’s valuation is a bit wild—$157B on 13.5x forward revenue? That’s Meta/Facebook-level multiples at IPO, and OpenAI’s economics don’t scale the same way. Generative AI costs grow with usage, and compute isn’t getting cheaper fast enough to balance that out. Throw in the $6B+ infrastructure spend for 2025, and yeah, there’s a lot of financial risk. But that said... their growth is still insane. $300M monthly revenue by late 2023? That’s the kind of user adoption that others dream about, even if the profits aren’t there yet.

Now, the “no moat” argument... sure, DeepSeek showed us what’s possible on a budget, but let’s not pretend OpenAI is standing still. These open-source innovations (DeepSeek included) still build on years of foundational work by OpenAI, Google, and Meta. And while open models are narrowing the gap, it’s the ecosystem that wins long-term. Think Linux vs. proprietary Unix. OpenAI is like Microsoft here—if they play it right, they don’t need to have the best models; they need to be the default toolset for businesses and developers. (Also, let’s not forget how hard it is to maintain consistency and reliability at OpenAI’s scale—DeepSeek isn’t running 10M paying users yet.)

That said... I get the doubts. If your competitors can offer “good enough” models for free or dirt cheap, how do you justify charging $44/month (or whatever)? The killer app for AI might not even look like ChatGPT—Cursor, for example, has been far more useful for me at work. OpenAI needs to think beyond just being a platform or consumer product and figure out how to integrate AI into industry workflows in a way that really adds value. Otherwise, someone else will take that pie.

One thing OpenAI could do better? Focus on edge AI or lightweight models. DeepSeek already showed us that efficient, local models can challenge the hyperscaler approach. Why not explore something like “ChatGPT Lite” for mobile devices or edge environments? This could open new markets, especially in areas where high latency or data privacy is a concern.

Finally... the open-source thing. OpenAI’s “open” branding feels increasingly ironic, and it’s creating a trust gap. What if they flipped the script and started contributing more to the open-source ecosystem? It might look counterintuitive, but being seen as a collaborator could soften some of the backlash and even boost adoption indirectly.

OpenAI is still the frontrunner, but the path ahead isn’t clear-cut. They need to address their cost structure, competition from open models, and what comes after ChatGPT. If they don’t adapt quickly, they risk becoming Yahoo in a Google world. But if they pivot smartly—edge AI, better B2B integrations, maybe even some open-source goodwill—they still have the potential to lead this space.

And then there’s the whole repetition issue. Infinite loops with "Pygame’s Pygame’s Pygame’s" kind of defeats the point of quantization if you ask me. Sure, the authors have fixes like adjusting the KV cache or using min_p, but doesn’t that just patch a symptom rather than solve the actual problem? A fried model is still fried, even if it stops repeating itself.

On the flip side, I love that they’re making this accessible on Hugging Face... and the dynamic quantization approach is pretty brilliant. Using 1.58-bit for MoEs and leaving sensitive layers like down_proj at higher precision—super clever. Feels like they’re squeezing every last drop of juice out of the architecture, which is awesome for smaller teams who can’t afford OpenAI-scale hardware.

"accessible" still comes with an asterisk. Like, I get that shared memory architectures like a 192GB Mac Ultra are a big deal, but who’s dropping $6,000+ on that setup? For that price, I’d rather build a rig with used 3090s and get way more bang for my buck (though, yeah, it’d be a power hog). Cool tech—no doubt—but the practicality is still up for debate. Guess we'll see if the next-gen models can address some of these trade-offs.

Has anybody already been able to successfully use prompt jailbreaking or other tricks to overcome this? It would be interesting to see what DeepSeek actually knows instead of what it is responding.

Censoring a model via selective training data or post-training is much more difficult.

The possible "solutions" applied to this "problem" (in the eyes of the censors) will be of high importance moving forward.

Other gov. actors also have an interest in altering models, let's not forget.