But then I looked at emails last year and Amazon was sending item names after order again. So it seems like this practice has changed multiple times.

Another interpretation would this is a counterfactual savings, like they previously paid $1M for y tokens, and now that tokens are cheaper they increased usage and paid $1M for 20*y tokens.

> but you’re charged for the report itself separately from the summary, and you are not allowed to access the unsummarized report

This analogy works if the LLM provider promises you access to the reasoning tokens, and fails if they don’t.

Anyway, you can distinguish this from the debate over copyright.

Regarding Claude, it’s true they must be losing money on free users since they promised they won’t put ads there.

Speculatively, I think people are overestimating the cost of inference for the consumer free tier chatbots. I suspect it’s effectively a marketing cost. The primary expenses are compute to train the next model, stock compensation, and heavy-duty enterprise inference (which pays for itself).

1) refuting the Jacobian conjecture

2) keep repeating the same disproven statement, because your priors can’t be affected by new evidence

1) Someone in the administration is leaking news soon before it’s announced, enabling insider trading. This is obviously supported by the evidence from markets.

2) The administrating is scheduling world events as part of a larger conspiracy to do enabling insider trading on specific predefined dates. Markets moving a few minutes early doesn’t provide evidence that this is happening.

Statements like: “There is a reason why these ‘peace’ deals are always announced on Fridays” imply the second (and aren’t even objectively true).

If there’s no difference between self-interest and what’s best for everyone then the whole discussion is meaningless - why even bring up self-interest in that case.

In fact there are many cases where the interest of some individuals conflict with the greater good - eg. Jones Act, sugar quotas, upzoning popular urban neighborhoods, military base closures, coal plant closing, countless others. And anyway framing this as an issue that only affects liberals or conservatives is incorrect, all of those examples cut across partisan lines.

> argue that voting is irrational because its unlikely to have any impact

The post (I’m not the author) actually argues that voting can be rational if you care about the impact on others. The low probability of being the deciding vote is multiplied by the huge impact on the nation as a whole by the better candidate winning. If you only care about yourself, the low probability of vote matters multiplied by the impact on yourself yields an effect that’s too small to care about in expectation.

> If what you're voting for seems bad for some members of the group, but good for you, it just means you have imperfect knowledge of what's actually good for you in the long term.

An 58 year old worker at a military base may be genuinely correct that closing the base is bad for him personally in the short and long term, but that doesn’t mean it’s bad for the public overall.