I cancelled my AI subscriptions because I am worried about catastrophic risks

TheManxLoiner

This is a linkpost for https://lovkush.substack.com/p/i-cancelled-my-ai-subscriptions-because

[Today is Day 13 of an ongoing 30-day microblogging challenge. It was initiated by Zoe Isabel Senón. You can see who is participating and also join in via this doc.]

In my post about a conversation with a capabilities research, I ended with this thought:

I believe that people more broadly should have personal redlines regarding the use of frontier AI systems. There is something incoherent about believing a a company might cause human extinction, but then also being their customer. This is something I still need to think about for myself.

With some reflection and a nudge from a friend who is currently boycotting OpenAI, I decided to cancel my two AI subscriptions with Anthropic and Google. Anthropic asks for feedback as to why you cancel:

I believe advanced AI can cause global catastrophes, and am now boycotting the subscriptions and credits from frontier labs.

As of now, I am still going to be using AI services but only the free tiers. Over the coming days and months, I hope to become more deliberate about how and when I use AI, and who knows, maybe even drop the use of general-purpose AI’s altogether.

Is this ‘rational’? Does this even make a difference to catastrophic risks? I was spending ~£200 per year on each subscription, whereas investments in AI are on the order of £100,000,000,000. Google nor Anthropic are going to change their actions based on my cancellation. But it does feel more consistent with my current beliefs, and sense of what is right or wrong.

If you’re worried about catastrophic risks, I encourage you to consider such actions. However, I suspect my circumstances make the decision easier than for others:

I was not a power-user of AI tools, so the drop in utility from paid version to free version is minimal for me.
I am not working at the moment, or doing high stakes tasks, so any drop in productivity is hardly consequential.
By nature, I am not a maximizer and more of a satisficer.

If you are not ready to cancel or reduce your use of AI now, then I at least encourage you to create a regular reminder (e.g. every six months) to nudge yourself into thinking about this question.

Okay but are you going to do anything else to help? At least study how to effectively convey the arguments for concern?

Stopping paying a subscription seems less useful than doing almost anything else. It's easier than anything else, but it will also set you a little behind on everything you do - including helping with AI risk. The paid tier models are now meaningfully better at everything.

Yes, if/when I decide what to do to help, I will re-consider using paid models

I never gave any money to OpenAI or xAI (the most reckless large labs in my humble estimation) and do not intend to.

I've started a trial of using an open-weights model for most of my AI needs. If the trial is successful, I would still tend to be enriching and enlarging an inference-serving company and (indirectly) makers of hardware for inference, which is not good, but at least I would have stopped enriching and enlarging any of the labs except to the very minor extent of providing organic advertising to the lab that created the model I would be using.

There is something incoherent about believing a a company might cause human extinction, but then also being their customer.

Sad but rather not not incoherent at all for most realistically parameterized humans. I's a most common and mostly individually rational outcome of so many a Public Goods game like this one.

You can draw an exact parallel between this and someone boycotting cars and then riding the bus. That you will continue using free AI is such a damning statement. You could at least also ride a bike.

Why? Using free ai means the frontier lab pays for inference.

The downside of using free AI is that it makes their user numbers higher, which helps them raise more money. Historically, raising money was a more important source of funds than revenue for AI companies, although now they are starting to tilt toward profitability (and at this point investors care more about seeing revenue), so free user numbers probably don't matter as much in 2026.