It seems like the usual Wall Street consensus on NVIDIA got completely blindsided by... a blogger in Brooklyn and a Chinese AI company. Jeffrey Emanuel, working from his apartment, wrote a blog post questioning NVIDIA's valuation, and it *might* have actually contributed to a $600 billion drop in their market cap!

You have to admire the sheer audacity of it. Here's this guy, who used to be on Wall Street but now runs a side project transcribing YouTube videos, and he manages to see something that apparently like 61 out of 67 analysts with "buy" ratings on Nvidia completely missed. The article even mentions that these analysts often rely on "industry experts" – which, as Emanuel says, is "like asking the barber if you need a haircut."

And then there's DeepSeek, another little guy bringing NVIDIA's value down also. Two little guys taking on big guys.

The article even hints at a slightly wild theory: that NVIDIA employees in San Jose might have been reading Emanuel's blog and thinking, "Hmm, maybe he's onto something..." and started selling their stock options. Whether that's true or not, the idea of a blog post potentially influencing market behavor like that is pretty fascinating in itself.

Interesting (and not entirely surprising) data leak from DeepSeek. Between this and the inherent risk of any Chinese company having its data accessible by the CCP, it really highlights the need for users to take control of their AI interactions.

The good news is, DeepSeek can be run on your own hardware. For those concerned about data privacy and security (and frankly, who isn't?), running DeepSeek locally is a very viable option. The instructions aren't even that daunting, check out https://unsloth.ai/blog/deepseekr1-dynamic.

"Wrong side of history," Altman says, after locking down his AI garden and charging rent. Suddenly open source looks good when someone else is eating your lunch. Cry me a river, Sam. By making their model open source and free, DeepSeek out-"OpenAI-ed" Altman.

The rest of the world isn't shedding tears for OpenAI's lost lead. In fact, a lot of people are probably popping champagne corks. There's a global appetite to see the US tech giants humbled, and DeepSeek is serving up a delicious dish of comeuppance. Enjoy the taste of your own medicine, OpenAI.

And all those chip restrictions? Turns out constraints are a fantastic motivator. While the US is busy patting itself on the back for its tech dominance, China's been busy innovating around the obstacles. Maybe having everything handed to you on a silver platter isn't the recipe for long-term innovation after all. Who knew? I think China just moved from 12-18 months behind the US in AI technology to 3-6 months behind.

Oh, Silicon Valley panicked over $5.6 Million? Bless their hearts. That's like finding a twenty in your couch cushions and declaring yourself a billionaire.

DeepSeek's "miracle" is less miracle, more marketing. That $5.6M figure probably covers the electricity bill for the final training run, assuming they weren't running it on hamsters in wheels. It conveniently forgets the small detail of the billion yuan supercomputer they built before the export bans. You know, the one with the 10,000 A100s. Just a minor detail.

And their data? Let's just say if imitation is the sincerest form of flattery, OpenAI should be blushing. Remember DeepSeek R1's brief identity crisis as ChatGPT4? Classy. Distillation is a thing, folks, and public datasets are...public. Let's not pretend DeepSeek invented AI from scratch on a shoestring budget.

The real lesson here? Markets are easily manipulated by shiny numbers and breathless headlines. DeepSeek played the game brilliantly. But the idea that they built a ChatGPT-level model for the price of a used Tesla is pure fantasy. The true cost is hidden in the infrastructure, the R&D, and the likely "inspiration" they drew from others. But hey, $5.6M makes for a better headline, right? Pass the popcorn.

That proposed SEC rule on reporting stock lending to FINRA (second link in last paragraph above) is about naked short sales used to manipulate the market. It's not about investors borrowing against assets to avoid taxes.

Well, that is very good news indeed! All we need is one zero-cost threat to cancel out all the other threats!

Which Russia conspiracy turned out to be false? What are you even talking about?

Weren't all the 9/11 hijackers fully identified before they boarded? How will more identity tracking keep us safer? What exactly is the TSA supposedly protecting us from?

Isn't definition 3 pretty much the original meaning when the term originated in the 1930s(?)? Like, if you're black, you have to be aware of all the ways the system is stacked against you?

Is it funny because it's impossible to oppose, yet DeSantis opposes it anyway?

How can he be an "ex-founder"? Does he now no longer have founded the company, though he did have before? Can any time traveler help me out with these verb tenses?

Nobody's going to waste a zero-day on this kind of attack. They will use exploits that are years old and already well-known. It's only a risk for older or otherwise unpatched devices, which are actually very common in real life.

Bleet.

It is not.

Thirtyies ... Fiftyies

It's an attempt to make encryption failure seem more significant to people who think, "I have nothing to hide, why should I care?"