yxwvut

There's been a lot of discussion of the copyright questions surrounding things like Copilot, DallE, ChatGPT, etc. Is there any sort of measure of novelty in output for these generative systems? For example, it'd be interesting to see a secondary system that finds the most 'similar' examples from the training corpus (though, conveniently, that's often private).

Clearly, something that just spits out a training example as output for a prompt is less 'novel' than something that can integrate bits and pieces from a wide corpus into something new, but I'm not sure how quantifiable that is in practice.

I recently booked an international flight through google's 'Book With Google' feature. The flight, though listed as 'Economy', was booked as Basic Economy: unable to be changed, unable to reserve seats, limited carry-on, etc (with none of those restrictions listed at the time of booking).

Google has structured this system in such a way that they've absolved themselves of all accountability for accurate info and is refusing to assist or reimburse me (my travel plans have changed and I'm stuck with $3k in tickets I can't use). See the fine print here: https://support.google.com/travel/answer/7515668?hl=en&ref_topic=2475360.

Never book through this service - always through a carrier or real travel agency/service.