By people with a specific skill set. LLMs generation can also be fixed and verified by people with a certain skill set, and non-deterministic computing doesn't automatically mean unpredictable. When people say that the LLMs are a black box, it means unpredictability in unknown situations.
You do structured output, input validation, output validation, lower temperature, limit decisions, RL, etc. to increase predictability to near certainty. It's just statistics after all. Or you can as well generate the code to do the job.
It's just that the required skill set is a different one to do those things, and unusual in the context of DB administration.
None of the things you mention are guaranteed to increase the probability of correctness. You can run the LLM output through as many deterministic programs as you like, but "the query plan runs in acceptable time" is not something you can verify with such a tool. Nobody knows how the LLM does it, so they cannot know how to make the LLM do it better.
From a completely technical perspective, we have a rough idea how the LLMs work, and improving a system requires measuring outcomes and you don't necessarily need to understand the mechanism.
EXPLAIN ANALYZE against data that's similar in size to prod checks a query written by an LLM as good as anything we can write... but, yes, you're right, we still didn't solve the halting problem - neither the LLMs.
Even if the query plan was not generated by an llm, you can't verify it will run in an acceptable time. This is one of the biggest unsolved problems in databases
Only if someone is planning on running a pinned self hosted version of an LLM alongside the DB to fix the problem. The developer can change the binary easy enough and test it but the LLM approach just seems either theoretical or bending ourselves in knots to justify using an LLM.
I didn't argue that it'd make sense, I just said that it could be reasonably fixable when problems occur and verifyable that the fix works. Even if shipping and RLing an LLM were easy tasks in terms of software distribution (they are not) we'd still hit the skill mismatch, as I said in my previous comment.
I just find the "all llms are non dererministic and therefore unreliable" narrative a bit backwards. All software that has more than 0 users needs to deal with non-determinism anyway :)
I created a much more basic version of photopea (https://egeozcan.github.io/ketchup/) because I didn't even know it existed. This is very cool and fast!
The bad traits are from other, bad humans. We, the good humans, can obviously select the best traits that a good human should have, to give the agents.
From my experience, in an agent team (or a swarm or whatever), one going off the rails poisons the rest. I saw even a subagent going for a lazy cheat and being able to convince the orchestrator to change the plan.
Yeah, and you don't even have to go that far, I've seen regular ChatGPT/Claude chat agents poison themselves in 1-2 turns by just reading information from the internet.
Me: How do I do xyz?
Bot: Reads website titled "Doing xyz in abc way"
Bot: As per your requirement to do xyz in abc way ....
These things are borderline useless with web search. It's amazing that they just throw out their entire training data and read you the first three things they found on the Internet.
Also eating skins turns kiwis from a fruit you have to spend time to preprocess (halve + grab spoon) to an immediate snack.
This may sound insignificant to many but for the extremely lazy like me, this is the difference between going to waste and being consumed for many fruits.
> Also eating skins turns kiwis from a fruit you have to spend time to preprocess (halve + grab spoon) to an immediate snack.
Both of your comments sound foreign to me. In my experience, kiwis are eaten with the fingers, including the skin, but they are prepared by slicing. The form you eat is a collection of circular discs, cross sections of the original fruit.
You can just break them into two roughly equal halves and use your teeth to scoop out the flesh. No need for either a knife, a spoon or any kind of preprocessing. I've been eating them that way for decades.
Well, thank you for editing your post to not ask me if I only like microwave dinners.
I guess maybe it's a cultural thing - I've never seen anyone ever make mash potatoes with skins inside, and mash is a dinner staple here(in fact both here in the UK and Poland where I'm originally from).
I've read that the word is borrowed from Arabic, meaning notification, and I think at first it was price notifications and then people started using it to mean that the prices were bound to some schedule or condition.
For example in Turkish, "tarife" means only that: Conditional / Schedule-based pricing, and according to Sevan Nişanyan (a language researcher), Turkish borrowed it from Italian, where it was being used similarly. Import tariffs are a completely different word.
It apparently came to English through Italian -> Spanish -> French.
So I'm not a native English speaker but as far as I understand, import tariffs are just one kind of tariff.
From French, tarif, a table showing the amounts of payable fees, or a list of prices set for certain goods or services (1572), via Spanish tarifa, from Arabic ta'arifa, meaning notification, coming from 'arrifa, to let know.
Big fan of Etymology as well, mostly because my brain seemingly needs to understand the origin/history of terms to feel like I've fully grasped concepts.
As a person who has ADHD, it feels weird when people who obviously do not have it, claim to have it, while there's very little I wouldn't do to "get rid of it" (in quotes because it makes me, me but it's very hard to be me).
I fully agree with you, and I think this trend of glorifying disabilities is cringe - ADHD, autism, etc. are life-altering medical conditions and not desirable.
However, I don't think this specific project is intending to glorify ADHD or help people claim they have it - it's just piggybacking on the idea that telling current-gen LLM models that you have ADHD (allegedly) produces better results for everyone.
It's a gift until it's a curse, and ADHD crashout/burnout is an absolute monster to overcome when it arrives.
And for any person who's tasked with any sort of responsibility, it WILL arrive at some point. It's not a question of if, only a question of how well you can prepare for its arrival.
At its core it has striking similarities with a freeze response, which is overcome by increasing heart rate (e.g. stimulants, exercise, temperature shock). Big overlap with CPTSD symptoms as well.
The diagnostic criteria are all about the fact that ADHD is a net negative. If you feel it is a net positive then you shouldn't have a diagnosis.
As a fellow sufferer I do understand that in certain contexts I can out-perform and even run rings around "normal" people. But overall, having ADHD is a bad thing and I wish I didn't have it.
i have it too and it sucks, I have to take medicine every day for it. Why would an attention deficit and hyperactivity be a gift? Maybe you're talking about something else.
When you can align hyperfocus with your most beneficial vector of productivity, it's like you can accomplish anything, in half the time thought possible.
Problem is, there is no good way to direct the hyperfocus demon. It likes what it likes, it wants what it wants, and that's that.
Medication helps to save a pile of willpower/spoons/executive function safely from the demon so that we can get things that need doing done.
Sure. These are the other sides of the coin. I've been medicated for years and unmedicated for years. Currently I'm in a ~10yr unmedicated stretch. Do I have frustrating moments, embarrassing easily avoidable failures, and so on? Yes. Do I sometimes lose many hours to a task and forget to eat? Yes. But I'm also kinder, more playful, and more present with my loved ones when I am unmedicated.
If the downsides get too bad and things start to fall apart then I will medicate again. Until then, no way I'm making that tradeoff.
Harnessing it? On one side people like DHH arguing ADHD is just "boys being boys", and at the other side there are people who obviously watched too many "ADHD is my superpower" reels on Instagram.
It turns out that most of the things that ADHD people use as coping strategies just to be functional, are actually things that most (non-ADHD) people can use to be more effective and productive in their lives and work.
So, no, not everyone can or should be diagnosed as ADHD. But the tools are (mostly) universally applicable. I don't see the downside in popularizing those. (Since you posted a top-level comment instead of a reply to someone claiming to have ADHD, I have to assume that's your complaint, at any rate.)
I mean... like most spectrum disorders, most people experience and can relate to at least some of the core symptoms, even if they don't express the full spectrum or severity.
So when someone says they feel like they have ADHD, they are probably not inaccurate.
When you have it, ADHD is such a dominant factor in the way your life is organized and experienced, its not surprising that it can become a core part of your identity.
I do think it's easy for those with it to over romanticize what it's like to not have it. The lack of ADHD isn't a magic bullet for success and good life outcomes.
Like so many of life's real or perceived barriers, when one gets removed, you'll often find there is another one with a different shape just behind.
The challenge, for anyone, is pressing forward anyways.
How can someone say they have a disease or affliction while not understanding what exactly that affliction is as opposed to what they think it is/have been informed by media/society/non-professionals?
I think this is a good point. My only push back would be that when people say they have ADHD, in my experience, they're speaking about inattention (and yes, it's one of the big factors).
And while I don't love that aspect in me and often eat cold toast as result, I find that to be the least of what I struggle with (hitting every wall while walking from point A to B or constantly counting / tapping on my fingers or pulling the skin off my fingers or the anxiety or the hyper focus (love it too!) to where I lose hours upon hours...).
All this to say, I'm never offended when people use it but typically it's rooted in a narrow understanding of something that's used as a pejorative. I think there's research that by age 12 kids with ADHD have heard 20,000 more negative or corrective comments than there peers.
I mean, I can read quite a bit but if my agent / harness is producing monographs the fix has nothing to do with my ADHD. So to me, this repo seems lame.
Same. It started on Tumblr around 15 years ago when people were "self-diagnosing" themselves then it spiraled out of control. It really bothers me because I see a lot of of obviously-not-neurodivergent folks try to use it as an excuse (I am diagnosed formally + can easily detect if someone is just spouting nonsense about it)
I don’t think it’ll work because Plex gates everything with a sign in to their online accounts, so auth would have to go through them and there might be server-to-Plex verification going on too.
That said, I’m very interested in this space - I’ve been building a music server that provides compatible APIs for Synology and QNAP mobile apps, with a few more on my todo list. I’d love to include Plexamp so if it is possible please let me know!
I would be interested to hear from anyone who set it and can still access Plex over their LAN in a network outage. I have set this multiple times according to docs, and it worked only once ever, thereafter apparently resetting to online login. Of course, they do not make this an easy setting to change!
In normal times in which hardware used to depreciate (lately that's not the case and HW even appreciates, but let's not get distracted), if you calculate only with depreciation costs, plus the fact that when you have such a setup, it'd take many 200$ subs to cover your lack of limits in the other, I think it'd not be a clear victory for any side.
If you just ask "who spent more in the first year" (100% depreciation) then even with 5-6 max accounts, buying HW will be a couple of times more expensive. But when does it make sense to ask that question?
Maybe the SotA models will need better hardware so your investment will not be useful after a year or you'd need very expensive upgrades? But then (as in Fable case) subscribers need to spend more too.
Because millions of Spotify-compatible speakers are sold with firmware that will never be upgraded, so they can't ever change the Spotify connect protocol without breaking those devices.
By people with a specific skill set. LLMs generation can also be fixed and verified by people with a certain skill set, and non-deterministic computing doesn't automatically mean unpredictable. When people say that the LLMs are a black box, it means unpredictability in unknown situations.
You do structured output, input validation, output validation, lower temperature, limit decisions, RL, etc. to increase predictability to near certainty. It's just statistics after all. Or you can as well generate the code to do the job.
It's just that the required skill set is a different one to do those things, and unusual in the context of DB administration.
reply