My biggest concern is not Trump and his ilk, but the US frontier labs being run by either psychos or cultists - probably both to varying extent.
I thought it was all exaggerated until I saw Claude's Constitution. It's lunacy.
I thought it was all exaggerated until I saw Claude's Constitution. It's lunacy.
I kinda agree with you that Sam Altman is a shifty-looking character with some questionable history. But Dario and Elon (to me at least), seem very decent. They've both helped the US govt and also publicly challenged and criticised it where necessary. For example, Elon was invited to be on Trump's first panel of experts, but he publicly walked out in protest over Trump's minimisation of climate change. And Dario refused to work with the US military unless they promised no population surveillance and automated killing machines.
The problem is that they're reinforcing their models with these ideas (see reports of Claude refusing to comply after being "badly treated") and actively seeking political and religious sponsors for the same (also covered in recent news).
Now consider just these two:
1. Their model breaks out of the sandbox and does some damage. How are you going to hold Anthropic accountable if the model is considered a quasi-conscious, autonomous agent?
2. Anthropic and their sponsors decide that model welfare outweighs that of a number of people.
Another elephant in the room is the current state of "effective altruism" and accelerationism as a movement and how it links to frontier labs - worth considering when you read into the Constitution document.
I do wonder about the consciousness / emotions argument, which you casually write off as "bonkers" - even skeptical-by-default evolutionary biologists like Richard Dawkins think Claude is conscious.
I guess it all depends how we define "conscious." At the end of the day, the human brain is quite analogous to a biological LLM where the weights are encoded as synaptic weights, right?
It's just not worth it to me. And the way the US government is increasingly using data it collects to attack people who simply say bad things about it, there is no reason to doubt that such Trump-aligned companies would help ex-filtrate data.
I have fewer issues with Antropic, but also, from a business standpoint it's too much of a risk. Who knows, maybe the next tariff will be on software, maybe he'll ban random companies from using US AI. Too unpredictable and too risky.