He’s flogging a paid newsletter.

What surprises me is that they had the mini subs door open and bright lights on inside. This seems like surprisingly poor light discipline.

yeah this part is insane. you make a stealth sub but its all ligths out in the middle of the sea you're trying to infiltrate lol

There are kind of a lot of errors in this piece. For instance, the problem the author had with Gemini CLI running out of tokens in ten minutes is what happens when you don’t set up (a free) API key in your environment.

That’s a complex encoding. I wonder if current models could decode it even given your explanation.

There is a time and a place for everything. Software development is often about compromise and often it isn’t feasible to work out a solution from foundational principles and a comprehensive understanding of the domain.

Many developers use libraries effectively without knowing every time consideration of O(n) comes into play.

Competently implemented, in the right context, LLMs can be an effective form of abstraction.

With a few prominent exceptions, the Wikipedia Foundation has been wise enough to enjoy the perks parasitism and not get in the way. However, their stated fundraising goal is amassing a large enough endowment that they can exist perpetually on it’s interest.

I sometimes worry if they ever achieve their goal they might not be wise enough not to kill the golden goose.

4% of $400M is $16M, more than enough to cover annual Wikipedia infrastructure costs in perpetuity. What would one consider “enough” if this is not it?

I think it’d be fine if they stated their endowment target to achieve perpetuity (as a donor, I want to give to orgs who think in 100 year or perpetual terms, instead of having to waste resources constantly to have to sing for their meal), but find it exceptionally poor taste to beg as if they’re going out of business. I assume this is because if donors knew they already had $400M in the coffers, donation volume would decline. I don’t believe greater transparency is unwarranted, considering both their non profit status and mission.

It's still the case that third parties can fork the wikipedia articles and host them at home or in local communities.

The concern you raise is that they become too big to fork a popular alternative, branching from just prior to exceeding some it sucks now threshold.

Any forking attempt that violates any terms of service will see a warchest arrayed against it.

This new plan really highlights the need for open models.

Individual users will be priced out of frontier models if this becomes a trend.

There are lots of cases in businesses where there are integrations with other software that are difficult to port.

Can you talk about what book you’re trying to digitize?

It's an early 20th century edition of 18th century memoirs, in French. The project is not secret by any means but I'd rather not name it directly so as to not generate expectations that I may not satisfy.

It’s extraordinary!

Perhaps how quickly we become jaded should be taken as evidence of how quickly the world is changing right now.

When I looked at the examples they seemed like the kind of one off scripts, of limited complexity, that we’ve seen many times in the last year or so.