Why are we using Markdown? Why do I use it every day?

It's easy to write. It's easy to read. Despite the OP's complaints, quality parsers exist.

pandoc can turn it into almost any format. We will still be writing markdown in 50 years, because the design bridges a bunch of compromises very nicely.

This general attack pattern is: get a lender with good collateral to call your bad collateral good, then swap collaterals, and it's a known and bad attack vector; the ongoing tension between innovation / speed and caution continues.

There's probably a flash-loan multiplier angle here for an even worse attack; I'm imagining chaining a liquidity change in the trusted price oracle for the CVT token in the middle of the swapping. Anyway, upshot - don't loan against North Korean attack tokens. Put it on the list.

It's not a bail-out Elon wants, nor is it 'safety' for SX. He wants capital to go build out space as fast as possible; he needs retail capital to do it.

I also like SpaceX - one thing many of the kids around here seem to forget is that elon has managed extremely dire capital and earnings situations very ably in the past - the above list for Tesla ten years ago looked much much worse.

This isn’t dispositive to success on your list but it does mean you can treat the company more like a long call : it almost certainly won’t go away.

One thing I've turned up in smaller models and I'm sort of winding my way toward verifying in larger ones is that if you train the MoE model from scratch with this kind of knockout / subset of experts baked in, then you get significantly better loss outcomes. In small models, it's actually better than training an MOE without conditioning on a reduced set of experts per pass.

Anyway, pretty cool. There's some Pareto-optimal curve based on memory bandwidth, amount of GPU / unified RAM and inference compute times for streaming stuff in.

  > The suggestion engine (korb suggestion threshold) is re-implemented in Lean 4 with five mathematically proven properties: suggestions have positive frequency, are sorted descending, come from ordered and available products, exclude basket items, and respect the count limit.

Teotihuacan shows excellent astronomy, civil engineering, and a very large implied economy built so long ago the name of the people group is not recorded.

Did Europe rock the Americas, hard? Yes. Was it because they were more advanced? In wartime tech, and psyops, yes. The rest? I would be cautious.

I don't know what the OP does for work, but almost certainly it's going to be easier to just start over with a new website. Maybe get the daughter's laptop taken to a data recovery specialist and try and pull browser history for the thesis.

This would leave holders who did not sign in two categories:

1) If you never sent a tx with an address, then you did not reveal your public key, and have some safety, e.g. you could do the PQ signature, wait, and be fine.

2) If you did, then you revealed your public key, and didn't bother to make the cutoff, and well, too bad.

There was a bunch of frankly dumb analysis about how long this would take the chain to process and how expensive it would be assuming that miners would all continue to enforce 10 minute blocks and transaction fees for these signature txs. I would be very surprised if the mining industry shot itself in the foot like that. The actual time to process 200mm or so new signatures just isn't that long. Hey we could do it on Solana if we needed to. That said, I imagine the papers this week plus Google moving up its timeline mean that there will be a concerted effort in Bitcoin land to get a real process down and tested in the next couple of years. Pretty cool.

Finally, I've read very little analysis about whether or not miners would choose to continue the energy dependent nature of mining, or try and move on. I think this is a pretty interesting economic question; I'm looking forward to finding out the answer. I expect mining will have a longer lead time than the signature problem - we're a long way from having Grover implementing SHA-256 as far as I know. And even then you still have 128 bits to deal with ONCE you get an equivalent amount of Grover-capable quantum compute out to the current ASIC ecosystem.