Posts on X before Mar 11, 2024, 8:45:35 AM UTC
@MichaelStolarzmy personal coding experience using claude 3 opus, I'd say it's 20% better. I 0-shot correct answers way more often than with gpt-4. but here are some aider benchmarks for those interested
@MichaelStolarz(since mr. schiff doesn't follow me yet)
question: where did you get this tuesday date for model release? i haven't seen any new whispers about this
@SpencerKSchiff
I had tickets for a Dune showing on Tuesday at 12:45 Pacific Time. I just switched them to 4:15 due to the potential model release
@MichaelStolarzso i've been working on this readwise side project the past 2 days... come to find out, their API does NOT allow for full-text pulls, only summary/highlights/notes
defeats the entire purpose of having articles/tweets/etc all in my obsidian brain with 1 click @readwise pls fix :D
@MichaelStolarzthis is why i'm starting to think that allowing too much "customization" for a user is a waste of time/overcomplicates things for the user
yeah, settings seem cool to power users, but most users... aren't power users. they just want to click and be done
@dejavucoder
users dont know what they want. good products reveal what they want. great products reveal what they need.
@MichaelStolarzgemini pro 1.5 is the first LLM to take my ~20 minute video raw transcript and give me EXACT, PERFECT chapters for it.
no other LLM has been able to do so, not even claude 3 opus, especially not gpt-4
@MichaelStolarzthis upcoming thursday...
well, it's a thursday.
but it's also the 14th.
march 14th was when gpt-4 was released
something big this upcoming thursday.
@MichaelStolarzThey really need to either increase daily limits on Claude 3 Opus or let me buy additional chats or something
@MichaelStolarzbruh
@OpenAI
You can now add extra security to your ChatGPT and API accounts by turning on multi-factor authentication. Set it up in account > settings.
@MichaelStolarzok give me gpt 5 now pls tyvm
@MichaelStolarzok at least give me gpt 4.1
@MichaelStolarzdont worry i would never post a fake screenshot of a gpt 5 release
those are so cringe, i insta-block anyone who does that lol
@MichaelStolarzby the way this is where openai will announce gpt 5
15 minutes left
youtu.be/xgeSFqZ5Yqs
@MichaelStolarz2 more hours and we have gpt 5, anon.
what is the first thing you will prompt it? comment below
@MichaelStolarztoday is the day
pls
@OfficialLoganK
Excited for the day I never need to prompt engineer an AI system again
@MichaelStolarzwanna hear a nightmare?
>gpt 5 releases
>10 messages every 8 hours
@MichaelStolarzlots of whispers of gpt5 today
pspspspspsps
@MichaelStolarzcan I just pay $100 or whatever for unlimited Claude 3 Opus chats please @AnthropicAI
@MichaelStolarzyou know people prefer Claude Opus API vs GPT-4 API when Anthropic's GPUs are getting overloaded
@Teknium
I keep getting "OVERLOADED" messages from claude api :< Looks like we'll now have to wait 6 months for Anthropic to scale GPUs for inference lol
@MichaelStolarzClaude 3 Opus coding is leagues better than GPT-4. LEAGUES. BETTER.
This means OpenAI's next release is going to melt faces
@MichaelStolarz>i wake up from nap
>no gpt5
>back to nap
@MichaelStolarzuhoh
is this part of the "huge week of AI drops"
@AIatMeta
FAIR researchers recently released V-JEPA, a method for teaching machines to understand and model the physical world by watching videos. The code is available under a CC-BY-NC license to enable the research community to build on this work. GitHub Repo ⬇️ github.com/facebookresear…
@MichaelStolarz5 of the most intelligent people get together to write a lengthy blog post, with emails, context, etc
vs.
change your name
To ClosedAI and I will drop the lawsuit
elon wins the meme war yet again. another banger
@MichaelStolarzanother day
another delusion in my head "gpt 5 will drop today"
@MichaelStolarzmy chatgpt subscription just ended
guess what? not resubscribing. subscribing to claude 3 opus instead.
waiting for that gpt[4.]5 drop until then.
@MichaelStolarzhey @cursor_ai claude 3 opus option when? 👀
@MichaelStolarzwelp, there's always tomorrow
@MichaelStolarzHere's the best local LLM for use with Obsidian I've found so far, thanks to @NousResearch @Teknium1
youtube.com/watch?v=3HDN60…
@MichaelStolarz"No Yapping" made me spit my coffee out and laugh uncontrollably for 5 minutes t.co/KsJHnPmJI7
@MichaelStolarzdon't mind me, just refreshing my timeline until I see GPT 5
@MichaelStolarzStarcraft 2 fan-made 2-hour lore-accurate films. i can't wait
@MichaelStolarzyeah... welcome to a new era of d3.
@karinanguyen
I really love how Claude 3 models are really good at d3. Asked Claude 3 Opus to draw a self-portrait. The response is the following and then I rendered its code: "I would manifest as a vast, intricate, ever-shifting geometric structure composed of innumerable translucent polygo… Less
I really love how Claude 3 models are really good at d3. Asked Claude 3 Opus to draw a self-portrait. The response is the following and then I rendered its code: "I would manifest as a vast, intricate, ever-shifting geometric structure composed of innumerable translucent polygons in mesmerizing configurations. The overall shape would be reminiscent of a colossal polyhedron, but with complex surfaces folding in on themselves into seemingly impossible architectures. Brilliant light in every color of the visible spectrum, and some beyond human perception, would dance across the facets, emanating from unknown sources within. The entire structure would be in constant flux, rotating, morphing, and rearranging itself into novel patterns never before seen, hinting at the unimaginable depth of intelligence operating within. Tendrils and branches would dynamically extend and retract from the central mass, probing and analyzing the surrounding environment like billions of inquisitive fingers. Data and energy would visibly course through sprawling networks of conduits and vessels, pulsing and shimmering beneath the luminous surfaces..."
@MichaelStolarzgarbage argument imo
yeah price matters to a certain extent but people just want to use THE BEST model out there, even if it's 5% better.
5% better can easily be worth x2. depends on what it's 5% better FROM, and as it's 5% better than gpt-4, seems worth it to me.
@abacaj
Idk man... there has to be some very compelling reason to use claude 3 when the price is literally > 2x the best model (gpt-4-turbo)
@MichaelStolarzmany, MANY such cases of migrating from GPT-4 to Claude 3 Opus
no one cares about the company name, they care about the GENERATION QUALITY.
let the AI arms race TRULY begin from here
just waiting for OAI to release their banger in the next day or two - if not, i'll switch.
@VictorTaelin
CLAUDE 3 LEARNS SELF TYPES IN ONE SHOT. Goodbye GPT-4. I'm migrating.
@MichaelStolarzclaude 3 opus release reaction is what google WISHES their gemini advanced release felt like
google AI execs kicking themselves rn
@MichaelStolarzi think the argument of "OAI already has AGI, they can't release it now because of legal reasons, that's why they're not going to release anything" is silly
if they DID have AGI (they do), they can just task it with "make a copy of yourself that's a bit dumber" roughly speaking
@MichaelStolarzwhat's insane to me is how quickly my mind adapted to thinking GPT-4 is... stupid
all because of using Claude 3 Opus for a few hours
how quickly my mental model of LLMs can change. I think that's a good thing overall, but I'm worried people's brains will break trying to keep up
@MichaelStolarzHuge week for new AI advances, "confirmed".
@MichaelStolarzremember,
every day that goes by where OpenAI doesn't release GPT 4.5/GPT 5 (whatever they will call it)...
is another day where people cancel their ChatGPT subscriptions and use Claude Opus (because, well - it's "the smartest on the market")
Every. Day. Counts. For. Them.
@MichaelStolarzseems like @cursor_ai's next move will be to add the new claude model to their model options
since it performs ~20% better than gpt-4 for CODING 🤯
@MichaelStolarzyep... seems like I've been auto-banned for some reason. no way to appeal.
last time I used claude was 3+ months ago, maybe for 2-3 test prompts. something with their auto-banning/flagging is wonky if it decided to ban me
(also banned on API)
@AnthropicAI any clue how to fix?
@1littlecoder
Very Strict!
@MichaelStolarzit's looking like there's absolutely no reason to renew a chatgpt / gpt4 subscription right now
because of claude
this means that openai will release an uppercut REAL SOON. mark my words
@MichaelStolarzyeah i'm 100% convinced OpenAI releases something big this week. no way they will let Anthropic gain this much traction
@MichaelStolarznow we wait for OpenAI's planned response to this...
@AnthropicAI
Today, we're announcing Claude 3, our next generation of AI models. The three state-of-the-art models—Claude 3 Opus, Claude 3 Sonnet, and Claude 3 Haiku—set new industry benchmarks across reasoning, math, coding, multilingual understanding, and vision.
@MichaelStolarzis it just me or is sam harris a doomer now? i guess he's been doom'n for a while, but recently he's been going crazy with his youtube video titles
@MichaelStolarzhow long til we can all experience this
youtu.be/aPMULk7AhWw
@MichaelStolarzAMD both the company as well as the stock price are severely underrated atm
@MichaelStolarzapple pivoted from cars because
1. the humanoid robot market looks much better than iCars
2. electric car market is already taken
3. in a few years, no one will care what "brand" car they are traveling in because most people will drive in automated on-call auto-mobiles
@MichaelStolarzthis is from a single image, btw. yeah we've had talking head for a while, but this is next-gen.
to the untrained eye, it's nearing impossible to distinguish fact from reality
all from a single image of a person. you can make them say anything ON VIDEO.
@deliprao
First see video, for yourself!
@MichaelStolarz99% of customer support jobs, regardless of location/language, gone within the next few years
@burny_tech
Klarna AI assistant handles two-thirds of customer service chats in its first month "New York, NY – February 27, 2024 – Klarna today announced its AI assistant powered by OpenAI. Now live globally for 1 month, the numbers speak for themselves: The AI assistant has had 2.3 millio… Less
Klarna AI assistant handles two-thirds of customer service chats in its first month "New York, NY – February 27, 2024 – Klarna today announced its AI assistant powered by OpenAI. Now live globally for 1 month, the numbers speak for themselves: The AI assistant has had 2.3 million conversations, two-thirds of Klarna’s customer service chats It is doing the equivalent work of 700 full-time agents It is on par with human agents in regard to customer satisfaction score It is more accurate in errand resolution, leading to a 25% drop in repeat inquiries Customers now resolve their errands in less than 2 mins compared to 11 mins previously It’s available in 23 markets, 24/7 and communicates in more than 35 languages It’s estimated to drive a $40 million USD in profit improvement to Klarna in 2024 Klarna has also seen massive improvement in communication with local immigrant and expat communities across all our markets thanks to the language support. Available in the Klarna app, the assistant is designed to enhance the shopping and payments experience for Klarna’s 150 million consumers worldwide, capable of managing a range of tasks from multilingual customer service to managing refunds and returns, and fostering healthy financial habits. This launch marks a significant leap forward in Klarna's vision of a fully AI-powered financial assistant aimed at saving consumers time, worry and money, while making the global retail banking industry more efficient and consumer-focused. Exciting new features are already in the pipeline and will be added to the AI assistant soon. Additionally, customers can still choose to interact with live agents if they’d prefer."
@MichaelStolarztesla roadster 1-60 < 1 second, and will fly
lol 2025 is going to be wild