timClicks

It means something, because it an iterative workflow. If you're willing to burn tokens, it's possible for weaker models to implement tasks by incrementally improving drafts.

Subscriptions are a mechanism to attract developers, who then advocate that their company should use the API.

Facebook used the same strategy to compete with Google Maps. The company became one of the biggest contributors to OpenStreetMap, which everyone has benefited from downstream.

Submitting projects like this to HN is a good way to get them nerfed.

If they're optimizing for SVG generation, then that's an excellent outcome in my opinion. Vector images shouldn't be "pretty niche".

It's an interesting example of the metric becoming a target. Simon started drawing pelicans because it wasn't something that existed before.

Another format that's worth investigating is Asciidoc. It supports the richness of Docbook XML but has fewer quirks than rST in my eyes.

Sorry if this is somewhat pedantic, but I believe that only US companies (and possibly only Delaware corporations?) are bound by the requirement to maximize shareholder value and then only by case law rather than statue. Other jurisdictions allow the directors more discretion, or place more weight on the company's constitution/charter.

References only have a single bit available as a niche (the null byte), which Option makes use of for null pointer optimization (https://doc.rust-lang.org/std/option/index.html#representati...).

In principle, you Rust could create something like std::num::NonZero and its corresponding sealed trait ZeroablePrimitive to mark that two bits are unused. But that doesn't exist yet as far as I know.

Asciidoc corresponds directly to DocBook XML. They're two formats with exactly the same semantics.

MCP is an example of "worse is better". Everyone knows that it's not very good, but it gets the job done.

It's been a while since I've played in the area, but is PCA still the go to method for dimensionality reduction?

This reminds me of when I provided some impressions of Erlang as a newcomer to their mailing list.

One of my suggestions was that they include hash tables, rather than rely on records (linked lists with named key). Got flamed as ignorant, and I've never emailed that mailing list again. A while later, they ended up adding hash tables to the language.

There are a few commercially valid strategies.

1. Goodwill and mindshare. If you're known as "the best" or "the most innovative", then you'll attract customers.

2. Talent acquisition. Smart people like working with smart people.

3. Becoming the standard. If your technology becomes widely adopted, and you've been using it the longest, then you're suddenly be the best placed in your industry to make use of the technology while everyone retools.

4. Deception. Sometimes you publish work that's "old" internally but is still state of the art. This provides your competition with a false sense of where your research actually is.

5. Freeride on others' work. Maybe experimenting with extending an idea is too expensive/risky to fund internally? Perhaps a wave of startups will try. Acquire one of them that actually makes it work.

6. Undercut the market leader. If your industry has a clear market leader, the others can use open source to cooperate to erode that leadership position.

That's a curious remark, although I guess it doesn't look high level from the eyes of someone looking at programming languages today.

C has always been classed as a high level language since its inception. That term's meaning has shifted though. When C was created, it wasn't assembly (middle) or directly writing CPU op codes in binary/hex (low level).

Perhaps the most generous interpretation is that the authors were writing an article for people who do the naïve thing without reading the docs. There are quite a few people in that category.

> The secretive nature of Cyc has multiple causes. Lenat personally did not release the source code of his PhD project or EURISKO, remained unimpressed with open source, and disliked academia as much as academia disliked him.

One thing that's not mentioned here, but something that I took away from Wolfram's obituary of Lenat (https://writings.stephenwolfram.com/2023/09/remembering-doug...) was that Lenat was very easily distracted ("Could we somehow usefully connect [Wolfram|Alpha and the Wolfram Language] to CYC? ... But when I was at SXSW the next year Doug had something else he wanted to show me. It was a math education game.").

My armchair diagnosis is untreated ADHD. He might have had had discussing the internals of CYC on his todo list since its first prototype, but the draft was never ready.

When production took longer and costs were not optimized, there was time for editors and writers to spend time editing and writing.

Shipping highly optimized assembler for a program made to work on computers with 4KB RAM as a ~100 MB PDF is quite the flex.

I must admit that while it's computationally quite wasteful, the web page does look quite neat.

More generally, using the prefix "Modern" haunts every product name that uses it. Technologies move fast and modern becomes antiquated very quickly.

My workaround is to use a computer monitor connected to a Linux box that I actually control.

I've always assumed that they charge you for publishing.

I consider Rust's Result<T, E> and Option<T> to be monads. Is this incorrect?

I love how deeply nerd-sniped you have been by this topic. It's wonderful to be able to observe your delight in solving this. Thank you for sharing.

That's the intent of the document - to reach people for guidance. The document under discussion is requesting comment from industry.

Sorry for the huge delays and also for missing the LinkedIn message.

Your support for creators is very much appreciated. There is now light at the end of the tunnel. More info here. https://news.ycombinator.com/item?id=41907594

I'm extremely sorry about this. It's been such a huge wait for everyone who supported me. Thank you so much for holding on so long.

I finally have time to complete the material and updates will start appearing in the next few weeks.

Also, I am more than happy to issue well warranted refunds.

No, because your training data is only an appropriation of the actual workload.

We justified the upfront cost by treating ours as a cheap second car, rather than as a very expensive bike. It's been an incredibly positive experience for our family.

Wait until you encounter his executable research paper about executable research papers.

The NAND gates video is probably the closest humanity will ever get to perfection though.