OpenAI wants to give us 5% of its success. It's a bad bargain.

Sign in<br>Subscribe

Link: OpenAI proposes handing Trump administration 5% stake, by Cristina Criddle in the Financial Times<br>In order to ward off backlash against AI and curry favor with the Trump administration, Sam Altman has floated the idea of giving 5% of OpenAI to a wealth fund that pays dividends to both the government and citizens — and that every leading AI vendor should do the same.<br>“Sam Altman, chief executive of the ChatGPT maker, has argued that giving the public a financial stake in the company is the best way to share the upside of AI and has suggested a stake of this size in early conversations with the administration, according to two people familiar with the talks.”<br>It’s transparently a way to align everyone with AI vendor profits. If the sector increases in value, the government and the voting population benefit. If it decreases in value … well, the government is incentivized to prevent that from happening. It also wouldn’t be without precedent: it’s modeled on the Alaska Permanent Fund, which does this with oil profits for Alaskan residents. Intel is also now 10% government-owned, and the administration has reversed course to be behind it since gaining that stake.<br>Would a government whose revenues are directly linked to the performance of a sector be likely to enact hard regulations on that sector? Perhaps not. It’s not a slam dunk, though: for example, the UK receives significant tax revenue on fossil fuels, but still promoted electric cars. There are lots of factors at play, and profit alignment isn’t necessarily outweighed by the effects of other harms. (See also: cigarettes, which are taxed but also tightly controlled as an addictive carcinogen.)<br>Meanwhile, Bernie Sanders has pushed for closer to 50% ownership through a sovereign wealth fund. At this much lower stake, Sam Altman’s proposal uses Sanders’s democratic socialist “share the wealth” language as a way to launder OpenAI’s profits through a thin veneer of good ethics.<br>What’s also interesting to me is that all of these arguments assume that AI is going to be an enormous driver of wealth and innovation — but what if it isn’t? It’s another great way to advertise the technology as something world-changing that everybody must get behind right now.<br>Even if AI turns out to be what the people heavily invested in its success say it will be, it doesn’t stand alone as a sea change innovation. The personal computer, the iPhone, word processors, and spreadsheets were pretty transformational technologies. Should there have been a wealth fund attached to each of those? What, exactly, makes AI different?<br>The answer is that it represents labor displacement: people will lose their jobs. And if that’s actually going to be the case, we need bigger, more structural safety nets and reforms. Dividends from 5% of a sector aren’t going to replace wages at scale — and are heavily dependent on valuations continuing to rise. This proposal ties the welfare of people who have lost their jobs to the success of the companies that drove those losses. The incentives are perverse.<br>We shouldn’t accept this proposal. Instead, we should push for stronger protections and stronger regulation. If a sector can’t succeed without real damage to working communities, then it must not be allowed to. And if these claims turn out not to be true, then it’s an empty gesture designed to add credibility to a self-interested science fiction view of the future.

Read more

I have a theory about AI fake news site The Editorial

I was taken in by a fake AI news site. I think it may be an attempt to poison LLMs with propaganda.

AI's costs are going through the roof - so businesses are telling LLMs to talk like cavemen

Me think this hilarious sign of times.

They tell us surveillance makes us safer. It undermines our democratic rights.

Civil society actors in nearly every region of the world now operate under the assumption that they are being surveilled. The result is a less democratic world for everyone.

A concrete tool to help newsrooms cover emergencies

A new site to help newsrooms cover disasters is refreshing in its concrete practicality. Wouldn't it be great if these existed for every aspect of running a newsroom?

Mastodon

Bluesky

Threads

Subscribe to new posts

Ben Werdmuller explores the intersection of technology, democracy, and society. Always independently published, reader-supported, and free to read.

Subscribe