This post was originally published on this site.
DeepSeek is a wake-up call for America. An alarm bell for the nation. We cannot allow another country’s technological advances to endanger and replace wide swaths of the U.S. workforce. That’s our job. Only we can do that.Â
As DeepSeek, China’s open-source large language learning model, surpasses ChatGPT on the Apple App Store, rocking Silicon Valley and upending the stock market, one thing has become clear: We must find a way to stop this new AI before it takes our current AI’s job of taking away the rest of our jobs.Â
How did we get here? It’s simple, really. Our best and brightest were gathered together by our richest and most ruthless in order to create something incredible: a brand-new way to put people out of work. Now another company has used our technology to build a more efficient product that threatens to put our work (the work that puts other people out of work) out of work. Does that make sense? This AI trained itself on our work — which shouldn’t be allowed unless we’re the ones doing it. Unfair is an understatement.Â
Breaking news: As you’ve been reading this, Qwen has defeated DeepSeek. That’s right, Alibaba just unveiled an LLM that can put DeepSeek out of its job, a job that’s putting out of work our own AI, whose intent in the first place is to put everyone else out of work. And as you were deconstructing that last sentence, SkrAll has officially launched and unseated Qwen, throwing the tech world further into chaos. SkrAll, for those who are unaware, is an LLM that nobody else knows about, because I just made it up.Â
Beyond the obvious money grab, these rival companies are trying for something more sinister: They want your data. Our companies do, too, of course — but ours treat your data with respect. They make sure your data is protected and feels cared for. It’s not just a large language model, it’s a large language family. And sometimes, when your data becomes important, they introduce your data to the highest levels of government so the kind folks at the FBI can know exactly who’s responsible for such special data.Â
More breaking news: SkrAll has been replaced by our very own fewtir.0, which was trained on just a single piece of paper and already outperforms every other LLM in existence. It is also imaginary, but the hope is that the idea itself brings some investors back.
We should be horrified by the prospect of a cheaper, more popular product stealing wealth from hardworking, salt-of-the-earth American venture capitalists. Our corporate leaders deserve protections against this kind of attack. Do you remember all the complaints unions used to have about AI, back before we decided to just ignore them? Well, this time it’s real, because it’s happening to rich people. This is a complaint we must not ignore.
Look, I get it. We all want to innovate to the point of making humanity, as a whole, unnecessary. That’s something we’ve agreed is a net good: for nearly all of us to net disappear. What I don’t understand is how we could possibly stand for someone else making our role in that future superfluous. And by us, I mean the people at the top of the capitalist class — of which I am not a part, but would very much like to be one day. Here’s hoping they see this.