This is the free version of the Spyglass column, if this was forwarded to you, you can sign up here. If you'd like to read the rest of these posts and get them sent to your inbox once a day (on weekdays), you can upgrade to the paid plan. Thank you for reading.
Two weeks after the DeepSeek bomb went off, now the fallout is clearing. As such, you can start to see the actual aftermath coming into view. As expected, it’s certainly less than the complete catastrophe for the current powers in AI that some feared and others hoped for. In fact, my viewpoint remains that I believe DeepSeek itself will matter far less than the mentality shift around AI that it ended up provoking in many. We’re seeing that play out in real time at the AI summit currently taking place in Paris right now. Both France and even the EU itself are both saying the right things about trying to compete in the space and even putting their money where their mouths are. Yes, the EU! They’ve taken a break from their regularly scheduled penalizing Big (American) Tech programming to tout the advantages of AI. Sure, they mainly want and aim for it to be their own, home-grown tech that rises to the occasion. But even just acknowledging the obvious while admitting that their previous regulation of tech was perhaps a bit overzealous is progress. At the same time, you can’t help but get a sense that this is all a bit opportunistic. DeepSeek is giving these countries cover to change their tones in the face of a new US administration that has promised to no longer take their regulation of American tech laying down. And the tariffs (at least the threat of them) are flying fast and furious at the moment. The fact that Vice President JD Vance is a speaker at this summit is one clear signal from the EU: we intend to play ball. Another clear sign: Sam Altman’s participation hot on the heels of the ‘Stargate’ announcement in the US. OpenAI would be happy to take Europe’s investment in AI infrastructure as well. They’ll be happy to take any and all investment, in fact.1 Really, what I keep coming back to is the quote from Rumi, the 13th century Persian poet: “What you seek is seeking you.” Over the past year or two, everyone has been looking for an AI narrative, and DeepSeek has now come along to give everyone what they want, it seems. Almost as if it was all inevitable. Beyond Europe, even Big Tech has been able to turn this moment to their advantage. After the initial Jevons paradox spin, the narrative has now turned towards DeepSeek being far less of a breakthrough and more of a natural progression of technology, at best, and something far more nefarious, at worst. Certainly, their cost claims were overblown, but now some are even questioning if the technological leaps themselves are as well. The distillation of OpenAI’s work is less impressive than starting from scratch, but it also perhaps showcased to many that working more in the open is the way to go. If nothing else, locking everything down doesn’t appear to be working all that well… Somewhat related, the real breakthrough of DeepSeek, the product, may end up being the notion that surfacing chain-of-thought work to end-users helps to instill a deeper level of trust in the outputs. And everyone from OpenAI on down is now scrambling to add more such visibility to their products. Meanwhile, on the bigger side of Big Tech™, CapEx spend continues unabated by the DeepSeek situation. Never mind the fact that much of it was probably already committed and hard, if not impossible, to pull back from even if the companies wanted to — all indications right now remain that they don’t want to. Except for maybe Microsoft, which was crying “uncle!” on OpenAI’s needs before DeepSeek happened. Luckily, Oracle was right there to step in (thanks, in large part, to Elon Musk, in an ironic twist). But really, SoftBank was just waiting for a time to load up the capital cannon. Talk about what you seek seeking you… Apple too has been seeking a narrative that not only was it okay not to spend tens — or hundreds — of billions of dollars on CapEx, but that it might be the prudent decision. Again, the rest of Big Tech clearly disagrees, but at least Wall Street is happy with Apple for the moment. If nothing else, it may showcase Apple had the right mentality in not trying to actually sell AI, but instead baking it into their devices and services as a feature. Trying to upsell the technology was already proving tricky for Google and Microsoft, and that perhaps just got even harder. Meta though, is welcoming DeepSeek with open arms because it validates their open source (read: open weight) approach with Llama. And, you know, wouldn’t everyone prefer to use such models coming out of the US versus, say, China? And this undercuts the likes of Anthropic (and to a lesser extent, OpenAI), which rely on selling their models via APIs. Amazon is pretty much in this camp too, though their agnostic AI in the cloud model is somewhat undercut by their massive investment in Anthropic. Still, that seems like a good hedge — as we may hear more about in a few weeks when the “Remarkable” new Alexa is (finally) shown off… Google, meanwhile, believes DeepSeek is the perfect showcase to highlight the fact that their models are actually more efficient than what is being touted about R1. OpenAI is quickly jumping on this train too. Again, it’s less that DeepSeek found some miraculous new scaling technology, they just continued the progress that was already happening. Of course, not mentioned in that is the fact that the continued plowing of money into new models with (naturally) diminishing returns was keeping the focus away from this other side of scaling... And related to that, China too has clearly benefited from DeepSeek in showcasing that they’re both not that far behind in AI and actually, the chip ban put in place by the US isn’t really working. In fact, there's a strong case to be made that the ban is what led to the “breakthroughs” of DeepSeek — because they had no choice but to use the technology they had access to more efficiently. And that also meant that reinforcement learning was put to the test in a more meaningful way, proving that it’s perhaps more ready for prime time than much of Big Tech had thought. Meanwhile, talking to both investors and startups, you get the sense that they believe DeepSeek is the best news ever. They can compete (and invest) without needing to break the bank paying Big Tech for their AI – or worse, NVIDIA directly (or indirectly) for their chips. Again, this was always going to happen, but it does feel like DeepSeek accelerated the discussion, and more importantly, the price cuts. I mean, even Lina Khan is trying to use and spin the DeepSeek story to her narrative advantage. Granted, she's reading it exactly wrong. But still... Really, looking over all of these and taking a step back: is anything all that different than it was before DeepSeek detonated two weeks ago? I mean, not really! Again, it’s just a mentality shift that has taken place across all of these sectors and companies. That’s undoubtedly a good thing — it’s always good to pause and revisit your strategy — but it’s not clear just how much will ultimately change in the long run. It's worth questioning that too! Is Europe going to win in AI now? Are more startups going to succeed? Is Big Tech going to spend less? Is OpenAI going to raise less? The answers here are obviously not going to be black and white, but my point is that you could certainly make the case that nothing much actually changed as a result of DeepSeek in the longer term. And again, that’s largely because it just really highlighted what was always going to happen anyway. What everyone was seeking here is what was already seeking them. Deep, I know. So again, I go back to the notion that DeepSeek itself matters less than what it inspires. Maybe what we really hope for here is that it simply sped up progress, even just a bit, clearing heads and roadblocks. While the notion of this being a "Sputnik moment" is now being played up as hyperbolic, perhaps it's better to simply read it literally. When the Russian satellite launched in late 1957, the US was already close to space. In fact, were it not for some trepidation and bureaucracy, many believe that we could have gotten there first. It was a kick in the ass, not a fundamental rethinking of everything. Sometimes, that’s all you need. It lit a fire that stayed lit through a trip to the moon. Back to work. ⭕ A couple recent columns for members of 'The Inner Ring' include a look at a new potential Apple product line that doesn't sound all that compelling on paper, but when you see it in action... And what some of Sam Altman's recent thoughts point to for both the future of OpenAI and AI in general in the wake of what else... DeepSeek. ( Sign Up | Upgrade)
|
