I don't want to imagine a world where juries were responsible for reading briefs and making decisions.

If you have been following this story for the past couple of years, this has been disastrous for Nike.

Law is certainly a less verifiable domain

Anything but a log scale would be misleading. When showing long term growth of something that has grown so drastically, it is imperative to use a log scale so that 10% gains look consistent over time. Especially in finance.

If your statistics or workload change, this approach is useless. The hints are generated being generated ahead of time, taking 95hrs to do so.

There's quite a big difference between "AI" and "pure neural networks." No, frontier chess (Alpha zero etc) is not "pure," because there's Monte Carlo tree search and a hard-coded game rules implementation.

All of this was public knowledge, and there is nothing damning about knowing someone who worked on something.

Is there any "local"/minio-esque object store properly implements the full conditional GET/PUT/DELETE spec, ETag, etc.? I have struggled to find this in the past.

I implemented this (HS5). Etags match the AWS S3 ones as well.

Awesome, I will take a look! We need something for our end-to-end tests at work that we can use and trivially spin up buckets for millions of tests in parallel.

The incentives to create fake engagement to satisfy their actual users who want discussion are quite clear. Doing this also gives reddit an advantage when monetizing their data. Scrape posts & comments for free, and you won't know which are LLM-generated and which at least have a human in the loop (but the human still may use an LLM, so this is where fingerprinting comes in). Perfect for incentivising the ai labs / other consumers of reddit data to pay up for an official feed.

It seems to me that this is the way the market is heading. Strange bleeding edge perf or sprawling complexity feels like the last niche of front end engineers. The rest will move further down the stack or be replaced.