Actions Speak Louder Than Words
Talk is cheap, words are air, but actions speak true. What does the AI industry's actions say?
The new main character in the AI discourse is Jacob Coxon who claims that AI could kill us all in the near future. How, why? No one seems sure, and more to the point, no one seems to care. I don’t want to rehash the debate over how plausible such claims are (not at all), but rather highlight the ways in which Coxon’s statements run counter to his actions. Until recently he was employed at Anthropic, and before that OpenAI. If he really was afraid that the computer algorithm would become not just sapient, but evil, a Molotov cocktail would do more to save humanity than a press tour.
I don’t encourage sabotage or violence, but one has to ask why all these people who are so afraid of AI never resort to or even muse about destroying servers or deleting code. Their response would no doubt be that AI has such potential as a force for good that it would be a terrible sin to destroy it. But you have to question if AI is destined to become fully conscious and so intelligent and capable as to cure cancer and develop bioweapons, at what point does a sword of Damocles hanging above a data centre become a necessary precaution?
To be fair to Coxon, all his friends in the industry are equally bad at speaking through action. If AI is as capable as Sam Altman says, why does he not use it to do the things he claims it can? If a billion dollar company staffed with a single person is possible, maybe he should be that person and start that company as a subsidiary of OpenAI? If Dario Amodei’s claims that AI will cure cancer are true, why isn’t there a pharmaceutical department at Anthropic?
Real innovative companies often branch out as they realise how good their innovations are and where they can be used. SKF, which invented ball bearings, started Volvo to create a customer and a display of what its products could do. Doing so is not just good in the sense of creating a customer base but also a way to capture the profits of a downstream market. But AI companies are desperate for other companies to use their models, always shouting from the rooftops about how powerful they are, when you’d expect them to guard the models closely if they were as useful as they say. Surely it is more profitable to just make a cure for cancer than to license your software to AstraZeneca or Johnson & Johnson who in turn creates a cure for cancer and gets a patent for that drug, in effect guaranteeing them a twenty year monopoly on the literal cure for cancer.
The promise of AI is superior intelligence in a way which has few precedents. The best comparison I've managed to come up with is this: imagine you had modern computer with access to Wikipedia, Excel, etc. a hundred years ago. If your goal was to get rich would you:
a) Use your vast advantage in knowledge and technology to start a business which you could run better than your rivals?
or
b) Rent out access to your computer?
Or to put it plainly: if AI does what its makers say it does, would they act the way they do?
If AI is as good as the frontier labs’ CEOs say it is at hacking, as every booster with a podcast practically screams, as every pseudo-intellectual on twitter and Substack writes, why haven’t Anthropic and OpenAI used it to hack the Iranian Revolutionary Guard Corps? The US is at war, and these companies claim to have invented tech which could help by disrupting the enemies’ systems and could prove invaluable in providing military intelligence, and they don’t take the obvious step of demonstrating their technology on a hostile military? Sure, the Iran war is deeply unpopular, but hacking the Iranians would presumably help end it sooner.
AI has a very bad reputation, people generally don’t like it, and Americans truly hate AI data centres. Anthropic and OpenAI need a win, something to convince the public that what they do is meaningful, important, and helps the average person. Instead, everyone in the industry is on a strange loop of raising concerns over the technology’s ostensible power to kill us all every few months, but never taking any real step towards preventing or slowing down development. Again, this is not how you would expect people who genuinely think they may doom humanity would act.
This dynamic manifests in AI labs’ business as well. If these companies were as profitable as they say they are, they would report their numbers according to widely accepted standards. They would be happy to raise money via selling shares to the wider public rather than staying private and relying on venture capital. Business leaders are not a bashful bunch who prefer to downplay their companies’ successes. Is it more likely that Altman and Amodei are simply different, or do they have rational reasons for why they act – and don’t act – the way they do?
The defining feature of the AI bubble is not the technology, the companies, or the big personalities, it is what we would call tolkningsföreträde in Swedish. Roughly interpretative precedence in English, it is the idea that a group has the right to interpret an event or term first, for example a union might have the right to interpret an agreement before the employer. The media has given the AI industry itself the interpretative precedence on all AI developments. If Altman and Amodei comments on a topic even tangentially related to AI their interpretation is beyond reproach, no matter how little sense it makes, no matter how little evidence they are able to marshal, no matter how much their interpretation might serve to benefit them financially. Taking a step back, and trying to interpret AI development without the guidance of noted liars like Elon Musk and Sam Altman, you start to wonder. If it did what they say it does, would they act the way they do?
If you liked this post you can read a previous post about nr 10 North here or the rest of my writings here. I also have a section for longer reads I call essays here, I particularly recommend my series called The Bird & The Technoking exploring Elon Musk's takeover of Twitter, and its political and cultural implications. It'd mean a lot to me if you recommended the blog to a friend or coworker. Come back next Monday for a new post!
I've always been interested in politics, economics, and the interplay between. The blog is a place for me to explore different ideas and concepts relating to economics or politics, be that national or international. The goal for the blog is to make you think; to provide new perspectives.
Written by Karl Johansson
Cover photo by Gizem Gökce from Pexels, edited by Karl Johansson
Kommentarer