Steven (@stevenssup) header
Steven (@stevenssup) profile picture

LLM,ASR,AI,and Tech.

Joined October 2022
Photo by Steven (@stevenssup)Photo by Steven (@stevenssup)Photo by Steven (@stevenssup)Photo by Steven (@stevenssup)Photo by Steven (@stevenssup)Photo by Steven (@stevenssup)Photo by Steven (@stevenssup)Photo by Steven (@stevenssup)Photo by Steven (@stevenssup)Photo by Steven (@stevenssup)
My Claude account just got banned and I resubscribed only two days ago. Now I've lost both my account and my money with zero recourse. Absolutely brutal.@AnthropicAI @claudeai @bcherny
1
64
LiveBench finally reflects real-world performance - Fable 5.1 and Astra are on top - Opus 5.5, not far behind, excels at design - Argon beats Terra, good at data analysis - Google beats open-source for the first time - DeepSeek is the top open-source model
79
30
7
495
39,995
This is a really good one. I just asked Codex and Claude to deploy on my Mac to do file transcription.
Meet NVIDIA Nemotron 3.5 Transcribe 🎙️ An open-weight, promptable 30B MoE model with ~4B active parameters that brings transcription in 40 languages, speaker attribution, translation, contextual prompting and audio understanding into one checkpoint. Watch it keep track of who said what when everyone’s talking at once. 1/ 🧵
3
37
Steven retweeted
Nemotron 3.5 Transcribe ranks #1 among open-weight models on the @huggingface Open ASR English leaderboard with a 5.33% average word error rate. Download the model → huggingface.co/nvidia/NVIDIA… Learderboard → huggingface.co/spaces/hf-aud… 2/
12
15
8
473
53,194
Today we’re releasing Underdog Bark 3.0, the #1 frontier speech model It's a tiny 163 MB download as accurate as NVIDIA’s Parakeet (15x bigger). It beats OpenAI's Whisper large-v3 (57x bigger) It's fast- transcribes 1 min of audio in <0.25s on a Macbook. On recorded meetings (AMI), Bark 3.0 gets 7.82% of words wrong. Parakeet gets 9.42%. OpenAI's Whisper large-v3 gets 13.63% error rate Bark runs on both your Laptop & Phone. It’s free and open source under CC-BY 4.0 Try private on-device meeting notes & dictation, powered by Bark, in Underdog today: underdog.ai/bark
251
278
67
4,515
243,089
Why did my 5-hour usage limit drop to zero all of a sudden? This doesn't make any sense. The second before, it was over 90%, and all of a sudden it dropped to zero. @OpenAI @thsottiaux
5
Hats off to the innovators
Turns out the future does have flying cars
5
Steven retweeted
Gemini 4 Argon is now rolling out globally in Search! Built to solve your hardest problems, Argon brings big improvements to tackling nuanced, multi-step tasks in AI Mode. It’s better at understanding your intent and following your instructions in back and forth conversations. Available starting today for Google AI Pro & Ultra subs, tap the ➕ icon to switch over to Argon and test it out.
9
22
3
178
14,802
Gemini 4 Argon is here. It’s available globally today for developers and enterprises in the paid tier of the Gemini API and the Gemini platform, for Ultra subscribers in Google Antigravity, and Ultra and Pro subscribers in the Gemini app and Search. Hope you all enjoy it!
1,537
2,451
1,215
28,298
6,018,893
What is this? It's only Tuesday.
Introducing Muse Wonder, our new family of frontier models with state of the art performance for personal and professional agentic work.
11
Replying to @thsottiaux
@thsottiaux Using Codex is frustratingly painful right now. Even with 50% of my weekly limit remaining, I had to use a banked reset just to keep it working.
2
1
93
Please remove the 5-hour usage limit so we can actually get work done. Consuming weekly usage faster is much better than being constantly blocked.
3
I have 3 back resets, and this 5-hour limit is really ticking me off. I can't get the work done; I have to wait another 5 hours to resume.@OpenAI @thsottiaux
4
Hmm, this is interesting.
You can now run Googlebook OS on your Mac, unofficially of course An open-source installer script lets Apple Silicon Mac owners test Google's new desktop OS inside a virtual machine ✅ Details - androidauthority.com/googleb…
8
First of all, GPT-6.1-Sol is such a disappointment. It is slow, it couldn't get the work done, and it eats the quota like crazy. Now I switched up to Claude again. I couldn't be more happier.
Replying to @antonpme @OpenAI
sad to see, hopefully we win your trust back soon 🤗
1
80
At the end of this month, I'm definitely going to cancel my subscription to ChatGPT. I feel like every usage reset is like gaslighting. The problem persists, and nothing fundamentally changes. Correct me if I'm wrong.
1
24
Steven retweeted
Claude now works inside Google Docs, Sheets, and Slides, and those files also open inside Claude. In Google Workspace, Claude sits in a sidebar next to your file, reads what you have open, and edits it in place. You can approve each edit before it lands.
850
4,290
2,070
54,216
13,427,837
Why do business networks like Fox Business and CNBC feature so many politicians on their shows? YouTube channels dedicated to these outlets are expected to focus primarily on business, markets, and finance, which is what viewers are tuning in for.
6
At this point, I feel Codex’s reset has gone from gaslighting straight to love bombing.
Over the next 28 days, each day we’ll either ship one thing that is a clear improvement and relevant for most codex/work users or ship a full reset. Let the improvements begin.
7
I am genuinely wondering if this gentleman is on a $500 plan or something. nitter.gq/stevenssup/status/2106…
GPT 6.1 Sol is all you need! It’s my go to on both work as well personally work According to arena you should get ~5x more usage as compared to Astra
6