One of the funnier things in recent AI news, was when, after years of moaning about just how dangerous their AI is, and threatening it would get moreso over time, Anthropic expressed surprised—shock, even—when the US Government stopped it from exporting its technology. If you believed the fine folks at Anthro, then you should've been entirely unsurprised. They've been consistently saying in ever-increasing apocalyptic language, that, among other nasty outcomes, their products are hastening the likely collapse of civilization itself. With their recent "hack prevention" model, they were giving people everywhere, nefarious and otherwise, an easier, more powerful skeleton key to all software. And this was according to the computer scientists, not just the marketers. They professed to be shocked when the feds came in and told them not to ship this abroad. If that's true, then their safetyism marketing had been an unbelieved-by-the-teller tale all along. Just like a whole lot of other "world-changing" and similar nonsense from motivated sellers of technology. In news that should've surprised no one, the safety-conscious and governmentally-uncooperative team at Anthro complied with sufficient government-mandated software changes to sell their latest dangerous computer programs across the globe. Just as they are attempting to get right with the Pentagon—never believe a business's conscientiousness until you see it cost them money. That's why I found it funny: as much as businesspeople can gain from "white paper" blog posts and appearances on philosophical podcasts, you can tell exactly what really matters to them by what they actually do.
What I called "safetyism marketing," borrowing Jonathan Haidt's term, Cal Newport calls "doom trolling." He's an appropriate smartie for the topic, and has a neatly written piece asking the AI companies to quit it. It's the first link below.
In another funny piece of the AI discourse, the pope, of all people, appears to be sucked in by the safetyism. There's been any amount of discourse about AI and what it means for soceity. But rather than wait and see, the American pope is out in front. His predecessors responded to the industrial revolution after quite a few decades. He's on it fast. It could be too soon. Most of the negatives he sees all around us have been present since the dawn of the computer age (technopoly). The worse of the risks he's concerned about are potential risks. It might be that the economy is tilted off its axis; it may be that computer people accrue more power; we might have to redefine what it means to be truly human in an age dominated by general artificial intelligence. On the other hand, this could be WorldCom II. But unlike Newport, whose been telling the companies to stop doom trolling and telling us to evaluate them on what their software actually do, the pope seems to take the companies seriously. As a vendor of software, I think this is probably not the wisest approach to take with software pitches.
Still, he's the pope, and his wisdom is worth parsing. The second link is Yuval Levin doing so in the electronic pages of the New Atlantis. He's a smartie worth reading.
A last thought: if you have any money in any investment market of any kind, then you'll have to have taken a wildly specific path not to have exposure to and gains from the AI boom. If you're like sixty per cent of US adults, then you have at least a retirement account with some index funds. You might not know it, but most of your gains in the last five years have come from AI-related stocks, from the chip makers to the data center operators to the large technology firms themselves. It might not be a bubble, but the stock market's rising tide bas been driven by a small fraction of tech companies. That link above, to the "has AI made money yet" site, is a sobering one. The sector has gotten and spent a giant pile of money on capital expenses: chips, data centers, salaries for nerds, etc. These investments are being made, like any financial investment, to generate future revenue and profits. While revenue is climbing nicely, profits are not only a long way off, some charts indicate they're getting further away, not closer. As investors, we might have reason to be concerned about that.
Ben Evans, another frequent citation in my links, is the right kind of business-technology smartie to read on this question. He's done the research on OpenAI and has a sobering essay. It's the third link below.
The major A.I. companies seem to be following a darker and weirder strategy: They like to solemnly describe the harms that their models will cause, while acting helpless to do anything about it.
Pope Leo’s wise but frustrating first missive on AI
OpenAI has some big questions. It doesn't have unique tech. It has a big user base, but with limited engagement and stickiness and no network effect. The incumbents have matched the tech and are leveraging their product and distribution. And a lot of the value and leverage will come from new experiences that haven’t been invented yet, and it can't invent all of those itself.