velominati

Am I the only one that thought this would be an AI-generated video with Bender, the alcoholic robot from Futurama? Could somebody please make that?

Typesafe has been quite about the underlying technology behind Jev. Given the speed and cost my hypothesis is that it doesn’t input tokens the way that LLMs do, ie iterating over every word and drawing the connections between each. That is an o(n^2) problem which is why LLMs are so expensive as they scale.

IP is easy. Tractionis hard.

Wow - Google does not even bother to show benchmarks of these models compared to the frontier and Chinese labs - only against previous versions. I'm not surprised. Having worked there for years it was amazing just how inwardly looking the company is.

It's never OK to physically attack someone like this. Full stop.

Separately; Sam's belief that "AI has to be democratized; power cannot be too concentrated." rings incredibly hollow. OpenAI has abandoned its open source roots. It is concentrating wealth - and thus power - into fewer hands. Not more.

“ It is named after one of its developers, Dick Pick.[2][3]” you can’t make this stuff up!

This sounds like an incredibly foolish expedition. They took a boat to sea without practicing basic seamanship procedures (person overboard drills), an inaccessible life raft, and no practice as a crew. That’s not adventurous, it’s reckless.

Money helps, but it's neither necessary nor sufficient.

I've been at a company that many considered to have excellent design and that received a lot of attention here (Heroku). There was a culture that permeated the company around beautiful design and what I would call "maker excellence". It was a place where people who liked building great things (both engineers and designers) felt at home.

Thanks for posting the PG wealth tax article. It is a highly flawed analysis.

First wealth appreciates over time. So if your wealth appreciates at 15% / year, and the govt. taxes it at 1%, the net effect is growth rate is slowed to 14%. With these assumptions, someone starting with $1mm in wealth ends up with $2.6 billion after 60 years!

Second - PG ignores that most wealth tax proposals have a high minimum wealth - in the $50mm range. So there is no early compounding of the tax. Adding this into the model, the wealthy founder ends up with $3.3 billion after 60 years.

With no wealth tax, this hypothetical founder ends up with $4.3 billion. So, yes the government has taxed a total of ~ 25% over 60 years, but the founder ends up quite wealthy.

Yes, that is the correct term.

Is it the longest known exposure ever, or longest ever with a beer can?

Plus of the 10% of the shares soldo, most were to institutions with an understanding that they are going to hold it "long term" (i.e. which may be just a few months). Part of the IPO road show is for the company to select investors that believe in the company and who will be "good" shareholders.

It is a misconception that when a company IPOs and the stock pops, that the delta is money left on the table. The vast majority of the shares are sold to institutional investors. There are explicit or implicit lock-up periods in place. And most of the pre-IPO investors will typically have a 6-month lockup. So if 10% of shares are sold in the IPO and 10% of those shares are actually trade-able, only 1% of the total stock is liquid. That can lead to very large price swings (i.e. the pop).