thornewolf

Boolean search is pretty expensive computationally. Maybe they are so far disconnected from that initial implementation that there is no semblance of support anymore?

Read up on HATEOAS - Hypermedia as the engine of application state. Long story short, don't use J's to store state, use html responses.

Testing best practices have the opposite philosophy for the most part. Avoid abstraction as much as possible. Do repeat yourself. Because a bug in tests is insidious, so you want to minimize that. One of the best ways to minimize bugs is to explicitly avoid abstraction.

The music and tech industries are undergoing large changes and we need to make decisions NOW to make sure the industry stays healthy. We are unionizing to make sure that us workers are part of the decision making process.

Not sure if you go back and reread old comment responses, but I have made a lot of QoL improvements to the game - https://www.turingjest.app/

I will be moving away from the confusing public lobby system in short order, but the game should be smooth to play with a couple of friends you can send the link to.

I'm always surprised that the Chinese room is considered an argument *against" understanding. It seems self evident to me that that exactly is what understand is.

Thanks for taking the time to provide screenshots. That is the public lobby list. I just auto spin up some lobbies for people to join if they don't want to create a game themselves.

There are some bugs here and there that can cause a game to "collapse" for lack of a better word and never get carried forward into the next state. I think this tends to happen when the server gets overloaded and restarts. I need to look into why this happens on occasion.

The total players figure counts the number of unique visitors that have been assigned a session_id. I think it's a bit optimistic since it counts any visitor at all. (excluding google crawler bots)

I have patched the active games counter. It was counting games that had been created but never started.

Thanks a ton for your bug reports, it's helping me harden the server so people can have reliable experiences going forward.

I have added the rules but I'm not sure what you mean by CSS. Do you not feel like the UI is usable? I tried to create something very utilitarian - sort of the same philosophy as HN in that regard.

I did discover some performance regressions after this post, but I feel that is expected when getting the HN hug. Hopefully it was still playable for you?

Added game rules to the website. Thanks for the feedback!

You are completely correct. I was lazy and didn't want to add another page to the site so I put the instructions in the comments. #1 TODO.

Swapping to a persistent db to keep game states.

Sorry everyone, I think the server crashed and auto-rebooted.

You might want to remove the word for spoiler's sake. It is the same for everyone, like Wordle

That's a tough problem when generating word embeddings. Bodies of text will rarely write "walking walk" or "the dancing dance" consequently leading naïve implementations to consider them unrelated.

The previous discussion of this situation (link needed) had some discussion on typhoid Mary. That previous thread was actually where I learned about the story.

If you are interested in doing this in an automated fashion, see my GitHub repo. It uses telegram as a frontend for gpt executing code on the backend.

https://github.com/thornewolf/gpt-3-execution

I'm one of those people who is obsessed with viewing or listening to content more quickly. I've managed to push to 2.8x on video and 3-3.5x on audio depending on content.

I have suspicious is is bash, or python. I think bash is good because you can express concepts very concisely giving fewer literal characters to mess up on. Python because the data set is so massive while still maintaining a lot of power-per-character.

Conversely, I have seen it underperform for verbose languages like Java, I think primarily due to windowing issues and that less meaningful code can be represented in one prediction window.

This sounds a lot like https://docs.windmill.dev/ another open source solution. Additionally, if I want to run the code locally I can just self host it.

CRUD apps are frontend wrappers for a couple SQL queries, yet they have driven persistent business value over the years.

I do recognize that the value of the CRUD app is re-representing data; where data is the thing of value.

Does this mean that prompts are the worthless SQL queries and the data that informa those prompts the real value drivers?

This is a post from Simon Wilson's blog. I suspect some meaning is lost on you since you are not familiar with Simon's past projects and the lack of any sort of introduction is jarring.

Reconsider this article as more of a brain dump that communicates what he is working on, than an article for the whole population.

Very fun so far! I have been having trouble deleting any existing val or writing to an exported val from an unexported val.

A specific example is one I made "@demo.newPaste", which seemed to fail to write to "exported @demo.pastes" until I also exported "@demo.newPaste". Is this maybe a mistake by me, a bug, or intended behavior?

Since gpt can occasionally get little bits of code wrong that ruin the result, having some prepackaged functionality can prove useful. If you want to play around with letting GPT execute code, feel free to checkout my github repo that uses telegram as a frontend for a code-executing GPT: https://github.com/thornewolf/gpt-3-execution

I did exactly that! See [1] for my implementation of a telegram bot that is able to write and execute python code.

[1] https://github.com/thornewolf/gpt-3-execution

Personally, I'd like to do both the data science and the engineering. I'd also like to be in a team where each member does both. I at least "feel" like these two skills are close enough spiritually to one another that you can find a team composed entirely of people that do both.

It's ok to be that guy, but in the future please phrase your corrections in less combative ways. The way you wrote your comment comes across as rude.

The AI seemed to have gotten stuck during my game with it. After waiting a few minutes I opted to resign.

How so? What I read here is "A possible explanation for this otherwise almost inexplicable behavior is the fact that I am a member of a group that has historically had widespread experiences exactly like this one"

Every "successful" black person I know has experiences similar to this one. They are over performing along some axis at some point in their career, yet some senior inexplicably seems to dislike them. And I am not exaggerating when I say "every".

Followed quantpy tutorials to implement my own black scholes and heston pricing models last year. Highly recommend.

"Poor sports, I've always struggled to understand people who'd partake in a foot race and then get upset because someone walked out of bounds to skip part of the race"

Simply because the context is hacking does not mean that performing additional hacking outside of the context of the competition is in the same spirit. Breaking the rules isn't hacking better than another team, it's breaking the rules.