I think I hold a similar view to you and have the same question so maybe what I’ve been thinking about might be useful to you. Everyone is upset about CSAM but when you talk about it, it’s only about deepfakes.
I don’t think we can avoid a world where people can generate CSAM easily, so we have to separate the discussion between being able to do that privately and grok being able to do it.
It makes sense to me that we don’t want widely used websites to contain images of CSAM that you can’t easily avoid, it’s simply repulsive to almost everyone and that’s almost certainly a human instinct, I don’t think it needs to be much more complicated than that.
In terms of generating CSAM privately or even sharing it with other people, I think this is a much more interesting discussion. I think at this point it is an open question on whether it is harmful. Could this replace the abuse that is happening to create some of the real content? Does the escalation argument hold water - will people be more likely to sexually assault children due to access to this material? I don’t think we know enough about pedophilia to answer these questions but given that I don’t think there is a way to stop generating this content in 2026 we really need to answer these questions before we decide to simply incarcerate everyone doing it.
Is that a useful thought experiment? Claude benefits you as an individual more than a coworker, but I find I hard to believe your use of Claude is more of a value add to the business than an additional coworker. Especially since that coworker will also have access to Claude.
In the past we also just raised the floor on productivity, do you think this will be different?
No that’s not true at all. Humans can deal with ambiguity and operate independently. Claude can’t do that. You’re trading one “problem” for an entirely different one in this hypothetical.
Isn't that what polishing 'the prompt' does? Refine the communication like an editor does for a publication? Only in this case it's instructions for how to get a transformer to mine an existing set of code to produce some sort of vaguely useful output.
The human factor adds knowledge of the why that refines the results. Not just any algorithm or a standard pattern that fits, but the correct solution for the correct question.
people talking as if communication overhead is bad. That overhead makes someone else able to substitute for you (or other one) when needs happen, and sometimes can discover concerns earlier.
So what happens if this wins after a very low number of plays? What if it won twice in a row? Would the plays be reset because it isn't representative anymore? Or should it be left up to give a different message?
I don't think that's true for European tourists, $37 per person is a lot and there are so many other cultural sites to see that paying twice that could definitely drop this off the list during a visit to Paris.
Makes total sense from a US point of view though, you are already committing to spend so much to get there.
