much harder to train longer context inputs

Everyone trains on queries from other models, it's called distillation

Well, given AI content cannot be copyrighted, haters can hate hate hate hate hate...

That's a little disingenuous. If i buy a printer and use the ink in the cartridge to reverse engineer a beautiful red, have i stolen something from the printer manufacturer? Especially if they lose business because they no longer have what distinguished them?

Clean room design is not new (or illegal), but it's always been a form of stealing

what's wrong with the current ide tools?

I feel like that's how you get Microsoft where each division has a gun pointed at the other division

that sounds like a pm problem

depressingly enough, things that work on small scale architectures often don't work at larger scales

Yep, most of what's remaining fails to scale. But it's still a very solid filter.

Sure, there are things that don't work on small scale and then work on large scale. But they're rare, and they sure are going to be expensive to find and validate.

much easier to raise money as a frontier lab

Panamax ship is 5000 teu (twenty foot shipping container equivalent)

I think you get about 4 MWh per TEU ( based on my 12V 100Ah battery)

so about 20 GWh

At 170Wh per kg (and ignoring the weight of the containers and any safety considerations), 20GWh of lithium battery would weigh 120,000 tons. This is a lot more than a typical Panamax DWT of 60,000 tons, which also needs to include the ship's fuel, provisions, crew, etc.

>the original Microsoft business model

From 1981

>Microsoft, which needed an operating system for the IBM Personal Computer,[9][10] hired Tim Paterson in May 1981 and bought 86-DOS 1.10 for US$25,000 that July

For people who are unclear what Palantir actually does

Palantir is a tech platform that consumes data from their clients in return for providing high level data-driven insights. They assign FDEs (or consultants) to really learn the details of a customers data. Foundry allows them to get single pane view of the data in an org and they actually have both the tech and engineering skills to do the dirty data cleaning jobs.

For an extravagant fee, you give them your data, they clean it for you, and then those same FDEs can tell you interesting things that you should have known, had you actually done proper data architecture in the first place.

Does it add value, yes. Is the value worth the fee? Like snowflake, they throw some very very good parties.