AI is All Talk
- Karl Johansson
- för 2 dagar sedan
- 3 min läsning
So many announcements, so few era defining technological breakthroughs.
Anthropic’s latest bluff is to imply that “recursive self-improvement” is right around the corner. The idea is that Anthropic’s AI is now so good at coding that it is plausible that it could soon write a better version of itself which would then write a better version of itself et cetera. It is a cool premise for a sci-fi story, but it is not real. If it was, Anthropic would not act the way it does now. Anthropic calls itself an AI lab, but it mostly seems to produce uncritical hype through announcements and press releases.
Anthropic regularly releases announcements designed to mislead people into thinking that AI is as capable as humans, and always tries to make it seem as though Anthropic is the down to earth, reasonable AI company compared to OpenAI or xAI. Is it reasonable to announce that you may soon make a discovery which changes everything? Wouldn’t it be better to save the announcement until the task is achieved? After all there is a risk that you will not manage it, especially when “it” has never been done before, as in the case of recursively self-improving AI.
Indeed, if you think you are on the cusp of recursive self-improvement why announce it at all? Why not just reap the rewards of this incredibly powerful new technology? Some might respond safety, after all there are legions of pundits on twitter who claim that superintelligent AI could be the end of humanity, so warning the world about it beforehand could be necessary if there is to be a Matrix-style war with the machines. I find such arguments complete humbug; AI runs on servers so you can just unplug the computer to stop it. Or to borrow a phrase: Claude! How many divisions has he got?
As I argued in “No Acoustic Guitars in Silicon Valley”, this incessant need to announce their triumphs, the constant attempt to convince you of AI’s importance is a clear sign of weakness; real G’s move in silence like lasagna. The timing of the “announcement” is auspicious (or suspicious) given that Anthropic is rumoured to want to go public soon. For a company that is ostensibly great at creating powerful agents which can do the work of white collar professionals Anthropic sure is fond of press releases when public demonstrations would seem more fitting.
AI has yet to steal my job, and seemingly no one else’s either, it has not disrupted any market, has not lead to any new types of software nor new challengers in established software niches like word processors or photo editors, it has not discovered any new physics, and it has not had an impact on productivity. But it has led to uncountable announcements, press releases, column inches. Talk is cheap. Anthropic and Dario Amodei, take my advice and wait with the next announcement until you have actually achieved something.
If you liked this post you can read a previous post about oil markets 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
Sources:
Kommentarer