rubyskills

Was it HFT or just processing large client orders? I'm having a hard time finding this out

Substring search for the word delve in the intro paragraph is all you need. :)

We already have a push system called Check 21. You scan an image of a check and it will send the money instantly to the bank instead of through the Fed. It created after 9/11, as money couldn't move when the Fed was frozen.

Zelle exists and you can scan a check that was written to you (check 21)

Next business day. How long it takes can also depend on the receiving bank

I do enjoy reading actual dialog coming from Kenny. :)

Reads like a content piece to push pgvector.

This is good feedback. I think this is more of a critique of my writing style than anything. I need to work on this!

I do admit, I am a pretty big fan boy of pgvector (and postgres). I think it's a much better solution than some of these other vector databases that are available. I would like to see it gain more popularity in the AI space.

I've been watching this github issue for awhile and almost didn't post this.

"As we witness a mass migration from Twitter to Threads"

Threads launched yesterday, way too early to state that as fact.

Agreed, though they did just get 50 million users. That is competing with openai adoption speeds!

This isn't quite accurate.

GPT3.5 is 4k tokens and has a 16k version GP4 is 8k and has a 32k version.

You are correct that this needs to account for both input and output. I suspect that when you feed chat gpt longer it prompts, it may try to use the 16k / 32k models when it makes sense.

Yes I'd like the details on this. My experience has been the opposite of you prompt it correctly, or it has the algorithm or data structure trained in its model already.

I was able to use a json-schema shard to streamline the generation of the JSON. I think the implementation is more elegant than what I've seen in other openai libraries. I'm really excited about functions and their potential! What has been your experience with them?

I like this guided NSDR session with Andrew Huberman. He pulls you out after 10 minutes in a nice way that gently reintroduces your CNS to reality.

https://youtu.be/AKGrmY8OSHM

It's also harder to stream JSON? Maybe I'm overthinking this.

It surely can do this, but you have to define what your taste is. Then you can ask, does this comment match my taste?

He is attempting it! I've been following his journey moving his apps out of the cloud and I just say it, I want to try some of his deployment tooling he's making as a result.

That's where fractal Fibonacci retracements come into play. :D ;)

Sarcasm if it wasn't clear

You should check out Andrew Huberman's videos about sleep on YouTube.

I found that strategies to get my core body temperature down towards the evening, combined with morning light exposure and exercising daily have made a huge difference for me. I'm tired at night and just want to go to sleep. I have a lot more tips, but most of it you can get from Huberman.

I really enjoyed "Learn C The Hard Way" by Zed Shaw. His style of learning resonates with me.

https://learncodethehardway.org/c/

I only take L Theanine at night as a sleep supplement. It works well for that!

Have you tried porting the problem into postgres? Not all big data problems can be solved this way but I was surprised what a postgres database could do with 40 million rows of data.

It's funny because you could argue that slack and discord replaced IRC. I wonder if we will have cycles or if commercial hosted solutions will always win this arms race.

I'd say this is more true for C vs C++. It has been my experience that the interop between C and languages like Ruby is much easier, the tooling is usually much better (just use ffi!), and the complexity is less (since that is what we're aiming for right?)

Haha, love it. Both of your passions melded together! (We met in Michigan at our mutual rust conf btw)

This is much easier to do in a centralized futures market. I can't imagine a rollback in stocks being easy or possible.

Exactly this. If you're a market maker, likely your trades impact your own trades too. As you accumulate a position, your average price is going up with it. Trades should not just roll back because one large hedge fund screwed up. Imagine being a retail trader with that expectation. Would be nice!

This video does a really good job analyzing this situation with the IF 16:8 strategy. It's quite interesting that this study is backing up his thoughts on it too.

https://www.youtube.com/watch?v=zC9_g01disc

This is very true. Placebo effect is extremely powerful.

Also there are some theories that spiking mTor less frequently is better. Some interesting research being doing on sirtuins and their role in repair of systems as well, especially in a fasted state.

Yes, not only that, but the patients were obese. I think 16:8 schedules can work well to help maintain a healthy metabolism, but if they're obese already, they'd need to completely reset their baseline (ie longer duration fasts) to achieve the results.

Original study:

https://www.nejm.org/doi/full/10.1056/NEJMoa2114833

139 patients over 12 months.

“These results indicate that caloric intake restriction explained most of the beneficial effects seen with the time-restricted eating regimen,” Dr. Weiss and his colleagues concluded.

I personally do time restricted feeding and notice health benefits so I'm a bit skeptical of this study as there are a ton of variables to consider.

I think original studies done in mice that showed metabolic improvement were based on a time window that if translated to human time would be closer to 4-5 day fasts.

There can be other benefits to time restricted feeding as well for long term health and I use it to help set my circadian rhythm.

Anyway I have a lot of random thoughts on this issue but wouldn't take this study to mean time restricted feeding "does nothing". Curious about other people's experiences / thoughts.