You may also want to add it to the vercel.app page.

There's a lot of pre-processing, experimentation and validation that went into this project. The training data collection and sanitization alone is a big undertaking.

As for the blog post itself, from the article:

> Note: This blog post is 100% written by me. No AI has been used whatsoever.

Put another way: You can ask the LLM yourself to do this project? Please do, share your prompt, I'd like to see it.

I will mention that there was a paper from 2017 on "Bounds on the Satisfiability Threshold for Power Law Distributed Random SAT" that constructed what looks very nearly like a phase diagram for satisfiable instances of SAT.

[0] https://arxiv.org/abs/1706.08431

Comments URL: https://news.ycombinator.com/item?id=48404433

Points: 2

# Comments: 1

Reference? The source was dedicated to the public domain in early 2018, which coincides with the release of the game [0].

> So the model that made this economically viable was the proprietary control model.

This is a complete fabrication.

> Anyway, just because a handful of games can exist on libre models (even given what I've said) that doesn't mean the industry can survive with mandatory libre requirements.

Making a living from open source software is hard, game or no. Making a living as a game developer is hard to begin with and many proprietary games are not commercially successful or viable.

My point was that the ecosystem is a lot more complex than your reductive analysis.

[0] https://github.com/jasonrohrer/OneLife/commits/master/no_cop...

Pretty dismissive, no?

Jason Rohrer puts many (most?) of his games in the public domain, including "One Hour, One Life" [0] [1]. As far as I know, his game is pretty successful, by indie standards.

Teeworlds was at one point accepting donations, I believe [2]. Solarus has a donation page [3].

I'm sure there are many more examples that span the spectrum of payment options and cover different permutations of being online or offline.

To me, the deeper question is what are you actually purchasing? The bytes? The convenience? A slice of server resources? Developers and artists time?

I'm happy to give money to projects that I use, especially if it creates less friction than trying to go outside of the payment method and if the project is libre/free. I'm willing to pay for proprietary content but I have little expectation about what kind of service they're providing, especially they fold.

If there's a libre/free option, I would much prefer to invest in it. If there's a proprietary option that is asking for resources, I'm much less prone to give since it's clearly a transactional relationship.

[0] https://onehouronelife.com/

[1] https://github.com/jasonrohrer/OneLife/blob/master/no_copyri...

[2] https://www.teeworlds.com/forum/viewtopic.php?id=446

[3] https://www.solarus-games.org/about/donate/

[0] https://www.oekobon.de/

Comments URL: https://news.ycombinator.com/item?id=48022751

Points: 1

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=47951171

Points: 2

# Comments: 1

https://archive.org/details/hokusaiimayoyhi00kats/page/5/mod...

https://archive.org/details/imayoykushikisev1kats/page/19/mo...

Are you sure you're remembering right?

Here's archive's list of Hokusai books:

https://archive.org/search?query=creator%3A%22Katsushika%2C+...

[0] https://github.com/Seeed-Projects/reBot-DevArm/blob/main/LIC...

Comments URL: https://news.ycombinator.com/item?id=47802392

Points: 5

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=47802389

Points: 4

# Comments: 1

Let's say you're a company that's providing an internet connection to a business. The company trusts you, so there's only compression of bits over the wire, not encryption, and you're aware of the compression scheme the company is using to send their bits to you. You're charging the company a premium for using the line you manage but you also lease the line, so it's in your interest to compress what they give you as best as possible so as to make a profit.

Say the companies compression scheme is imperfect. They have a Huffman coding of their (imperfect) model of tokens they send, call it q(x) (that is, they think token x shows up with probability q(x)). You've determined the true distribution, p(x) (token x shows up with actual probability p(x)).

The business has tokens that show up with probability p(x) but they encode them with lg(q(x)) bits, giving an average token bit size of:

    -\sum _ x p(x) lg(q(x))

    -\sum _ x p(x) lg(p(x))

    -\sum _ x p(x) lg(p(x)) - \sum _ x p(x) lg(q(x)) = -\sum _ x p(x) lg(p(x)/q(x))

To me, this is a much more intuitive explanation. I made a blog post about it [0], if anyone cares.

[0] https://mechaelephant.com/dev/Kullback-Leibler-Divergence.ht...

Spending on social security is about $680B, medicare is $480B, defense is at $410B [1].

Microsoft's valuation is approx. $2.8T [2], Google $3.8T [3], Amazon $2.3T [4], Facebook $1.6T [5] (Linux supply side is valued at approx. $8.8T).

The FBI employs roughly 38k people with about $10B in funding [7]. The CIA employs roughly 22k people with about $15B (?) in funding [8].

So, from that perspective, $10B is roughly .5% of yearly tax revenue (and about how much the FBI/CIA are funded) and estimated 50k people is about the size of the FBI and CIA combined.

[0] https://fiscaldata.treasury.gov/americas-finance-guide/gover...

[1] https://fiscaldata.treasury.gov/americas-finance-guide/feder...

[2] https://finance.yahoo.com/quote/MSFT/key-statistics/

[3] https://finance.yahoo.com/quote/GOOG/key-statistics/

[4] https://companiesmarketcap.com/amazon/marketcap/

[5] https://companiesmarketcap.com/meta-platforms/marketcap/

[6] https://www.hbs.edu/ris/Publication%20Files/24-038_51f8444f-...

[7] https://en.wikipedia.org/wiki/Federal_Bureau_of_Investigatio...

[8] https://en.wikipedia.org/wiki/Central_Intelligence_Agency

[0] https://www.crowdsupply.com/oddly-specific-objects/sensor-wa...

[1] https://www.crowdsupply.com/oddly-specific-objects/sensor-wa...

[0] https://en.wikipedia.org/wiki/Steins;Gate_(TV_series)