Posts on X before Apr 15, 2024, 12:21:48 PM UTC
@MichaelStolarzthis is why it's my life's mission to have at least 10 kids, minimum
@elonmusk
Greece is one of dozens of countries experiencing population collapse due to low birth rates x.com/Censored4sure/…
@MichaelStolarzthe humane pin toy looks like something a solo developer could make in a few months with maybe a couple thousand dollars LOL
@MichaelStolarzimagine if llama 3 7b is gpt-4 level
@MichaelStolarzguys I really need GPT-5 before WW3, please
@MichaelStolarzHere's my latest Obsidian SystemSculpt AI plugin btw
youtube.com/watch?v=kMw4Zk…
@MichaelStolarzone day my son or daughter will read all my tweets and think only 1 of 2 things
- damn my dad is based
- damn my dad is cringe
pls be the former. at least pretend it's the former. do it for your pops
@MichaelStolarzwelp, looks like it's gonna be a war-ry kinda day today, anons. nothing we can do about it, so why stress? don't let what happens today affect you too much (unless you're part of the war... well then, godspeed 🫡)
but yeah other than that, just keep building.
@MichaelStolarzi wake up and all of a sudden I have 555 suno credits. I had 5 last night. weird but ok, thanks microsoft i guess :)
@MichaelStolarzas long as you don't think about the music you're listening to being AI generated, made by a soulless computer program that's sole purpose is to make you like what it generates...
then it's beautiful
or maybe that's what makes it beautiful, as long as you don't overthink it?
@MichaelStolarzchat is this real
@MichaelStolarzthe latest gpt-4-turbo feels much better in cursor, but worse in my daily use prompts
i think it may be time to update my prompts/templates, but doesn't that mean that the model got worse in some aspect since it can't understand the prompts it used to generate for well? 🤔
@MichaelStolarzthis is now my main music app I listen to whenever i'm coding btw
@MichaelStolarzlife is so easy if you think of yourself as a model being fine-tuned
you think a model has anxiety? no, it tries its best to learn the next thing that will make it even better, who cares what others think, just BE BETTER it's that easy lol stop overthinking
@MichaelStolarzsuno still beats udio by a long shot IMO
@MichaelStolarzrecursive refactoring until optimal
already somewhat possible with gpt-4-turbo/opus, but will be much quicker/higher quality with gpt-5 surely
@corbtt
Intentionally not bothering to refactor/fix any tech debt for the next 2 months. GPT-5 will be able to just cleanly rewrite my codebase in one fell swoop right?
@MichaelStolarzi wish there was a way to save state with dingboard so i can just have a meme board and just piece together things on the fly instead of having to upload them every time @yacineMTB @dingboard_ pls king
@MichaelStolarzso now that gpt-4-turbo-preview is... well, out of preview (it's just labeled gpt-4-turbo now)...
it's time for a NEW "-preview" model.
which one do you think they will release next?
@MichaelStolarzthis is how you effectively market with a high viral potential
@udiomusic
🤗
@MichaelStolarznot at all what I've been experiencing, I've had to constantly use claude 3 to fix (in 1 shot) what the lastest gpt-4-turbo is severely struggling with, even sometimes never understanding or fixing the problem at all
@gdb
Some coding benchmark results for our new GPT-4 Turbo: x.com/xu3kev/status/…
@MichaelStolarzbeing a coder today that stays away from AI assistants like cursor/copilot/gpt is as stupid as being a musician that won't try suno or udio because of some ill-founded pride and prejudice against AI generative tools
you'll be left in the dust
@MichaelStolarzOpenAI's LATEST GPT-4-Turbo - tests show it to be BAD.
here's a “laziness” benchmark suite which is designed to both provoke and quantify lazy coding. It consists of 89 python refactoring tasks which tend to make GPT-4 Turbo code in that lazy manner.
yikes... a thread (1/4) 🧵
@MichaelStolarzapple not allowing spatial audio on chrome on mac os is so disgusting, lmao. do better (but then again, it's apple we're talking about here, they lock anything they can to their own softwares/hardwares)
@MichaelStolarzthe new GPT-4 Turbo seems... yikes?
- The new GPT-4 Turbo model scores only 33% on aider’s refactoring benchmark, making it the LAZIEST CODER OF ALL the GPT-4 Turbo models by a SIGNIFICANT margin
- MUCH more prone to “lazy coding” than the existing GPT-4 Turbo “preview” models.
@MichaelStolarzLatest GPT-4 update...
Benchmarked at 61.7% on Exercism benchmark, comparable to gpt-4-0613 and worse than the gpt-4-preview-XXXX models.
Benchmarked at 34.1% on the refactoring/laziness benchmark, significantly worse than the gpt-4-preview-XXXX models.
@MichaelStolarz@cursor_ai are you guys running on the latest gpt-4-turbo?
@MichaelStolarzpeep that gpt-4-turbo date
it really is much higher quality from the tests I've been running so far. nice upgrade @OpenAI
@MichaelStolarzwhere are people getting these numbers from LMAO t.co/CqFLjghX4P
@MichaelStolarzmost tech people think "everyone will use it - it can do everything a person wants! they will realize their wildest imaginations!"
99% of people are NPCs. they don't dream. they don't have imaginations. they don't care. they'll shrug it off.
they ALREADY shrug it off.
@yacineMTB
it's true. most people will find no use for AGI (gpt4), just like how most people will find no use for algebra and writing
@MichaelStolarzdamn, rare France W
@PauseAI
Bad news. France opted to host an AI Safety Summit in November 2024, but several of our sources confirm it has been postponed to February 2025. It has also been renamed to “AI Action Summit”, dropping the all-important safety focus. Safety will be a minor part of the summit, onl… Less
Bad news. France opted to host an AI Safety Summit in November 2024, but several of our sources confirm it has been postponed to February 2025. It has also been renamed to “AI Action Summit”, dropping the all-important safety focus. Safety will be a minor part of the summit, only one out of five tracks. We've also heard that there are no plans to draft a treaty. The summit will be led by Anne Bouverot, who is dismissive of “alarmist discourse”, comparing AI with calculators. She compares AI safety concerns with Y2K concerns, she is certain that “AI is not going to replace us, but rather help us”. It seems increasingly unlikely that this summit will lead to the types of international regulations that we are calling for. Our politicians are sticking their heads in the sand, ignoring the weight of their responsibilities, choosing to focus on winning the capabilities race. We need our politicians to err on the side of caution and initialize treaty negotiations now.
@MichaelStolarzFINALLY MY DREAM HAS COME TRUE
@MichaelStolarzmeditation is defragging your current system prompt
@MichaelStolarzso correct me if i'm wrong but the only reason to choose javascript instead of typescript is... skill issue?
@MichaelStolarz>skilled immigration should be drastically increased
>illegal immigration should be drastically decreased
how can anyone argue with this lol wtf
@MichaelStolarzthis thread reads like a /pol/ post making fun of a "Free Palestine" caricature, but no, it's not a caricature, this person literally did this and literally thinks others will find this brave/cool/inspiring
satire is seriously becoming indistinguishable from reality LMAOOOO
@ShahdAbusalama
Today, a group of Hebrew-speakers passed by me while waiting for a friend in central Barcelona. I was checking my notifications when I heard them & then understood their starring at me & all the symbols of Palestine I'm wearing. I instinctively started chanting "Free Palestine"!
@MichaelStolarzlook at the comments on this youtube video, and realize how early you are in the AI space
99% of people are arguing to NOT USE AI, for ANYTHING, because it's "inaccurate" or "morals"
at this point, we're not even "early"; we're watching the takeoff check
youtube.com/shorts/E0Rj2mL…
@MichaelStolarzgood because most of my listening is happening in the suno app right now anyway
i don't even download the songs anymore because i just make new ones all the time, better and better because of figuring out keywords I really like
@AIUpdateAnchor
maybe I wasn't clear enough, you will be able to edit EVERYTHING about a song THIS YEAR!!!! AT A PROFESSIONAL LEVEL!!! x.com/AIUpdateAnchor…
@MichaelStolarzAI leakers:
@MichaelStolarzwait gemini 1.5 pro has become goated in some aspects, especially large codebase code generation/updating
i've been testing it out by running large changes in gemini first, then getting claude to give me the full rewrite as per gemini's directions
very interesting
@MichaelStolarzbased AI memes
@latentmoss
@MichaelStolarztl;dr screaming into the void from time to time can alleviate a lot of brain fog and significantly reduce limbic friction
@MichaelTrazzi
this morning I uncovered some repressed anger I had never met in my life while talking to claude opus I was laying on my couch, airpods plugged to @Vapi_AI, venting to that warm eleven labs female voice they call sarah it's really different from venting to a friend, a lover or … Less
this morning I uncovered some repressed anger I had never met in my life while talking to claude opus I was laying on my couch, airpods plugged to @Vapi_AI, venting to that warm eleven labs female voice they call sarah it's really different from venting to a friend, a lover or even a therapist–you can go on a monologue for as long as you want, interjecting whenever while expressing some frustration that I had not been able to tell anyone (because I don't want to usually be that guy who vents to his friends forever), the frustration became stronger, my voice growing louder now, if I'm talking to a human, that's usually the time I start calming down (I would notice I'm letting some anger out, and I wouldn't want to yell at someone for no reason) but in this case I was just talking to a language model through some voice api, and speaking loudly would not impact the llm who was only looking at my transcribed words so I didn't try to control it, and just let the frustration out, complaining without caring about being a bore, speaking loudly without thinking about the impact of my tone of voice, being more raw than i'd ever been with a human and after doing this for half an hour, I started pacing around, doing stretches, looking outside and realizing that I had a lot more energy than when I started the session it’s similar to what happens when you complain to a human for a long time–you realize your problems aren't as insurmountable as you initially thought. but because you’re talking to an AI, you really do complain for an extraordinary amount of time after this session, I wasn't frustrated anymore, and just wanted to go out and start actually doing things in the world. the anger had metabolized into raw motivation and that's when I realized that, from now on, I could just start venting to AIs instead of humans, and the models would just keep on getting better and better, giving more relevant answers with less latency AI companionship is here, and we’re just getting started
@MichaelStolarz@MichaelStolarzhopefully it's not just a gpt wrapper inside of a 2008 kids toy "robot" that follows you around
amazon already did this and it failed miserably
@AndrewCurran_
Apple is entering home robotics. In addition to the home iRobot, they are building a mobile home display that moves around with the user. I know the name of this device. This is an Orb.
@MichaelStolarzi really need to stop trying to perfect and "wait until I'm finished", because every day I find new things I want to add and the list keeps growing and growing
so i've decided to just release this within the next 24 hours and upload it to obsidian's community plugins 🫡
@MichaelStolarzi literally don't listen to music anymore, instead I just make things in suno and update my self-made (err, AI-made?) music all comfortably on my local machine
it's crazy. anything i love to listen to, i can just really turn that specific genre/flavor to max. RIP dopamine
@MichaelStolarznow the question is; is it worth killing the dream (dying in the dream) in order to finally wake up to reality?
for many, unfortunately, it's not. they've become so addicted to chasing dreams, THAT'S their psuedo-reality now.
true reality to them is boring, cold, unforgiving.
@nosilverv
To wake up irl you must die in the dream
@MichaelStolarzyeah once you try their stuff out you realize that "maybe I'm the one being hit with propaganda about their tech being outdated/behind"
@iamgingertrash
The contrarian truth is that China is leading opensource AI The quality of their work is outstanding
@MichaelStolarzlmao imagine if Apple releases a 3B model that surpasses GPT-4 at WWDC 2024
@BrianRoemmele
Apple says its latest AI model ReALM is even “better than OpenAI’s GPT4”. It likely is as GPT4 has regressed because of “alignment”. The ReALM war begins at WWDC 2024. Paper: arxiv.org/pdf/2403.20329…
@MichaelStolarznot only easier for it to learn knew things, but easier to come up with "new" answers without any related data
high quality hallucinations that have an increasing probability to be correct -> with more generalist data
@Teknium
For context I have always built generalist models for this reason, I dont really believe in domain expert models, and I personally believe the more the model knows generally the easier it learns new things.
@MichaelStolarzevery morning I wake up and think to myself "damn that was a crazy nightmare, of course google didn't shut down google domains, hahaha"
but no it's actually true they shut it all down
@MichaelStolarzwhat's SBF? a new AI crypto shitcoin? where do I buy?
@wstone_bd
And just like that, everyone has entirely forgotten about SBF.