HALO terminal (@HALOterm) header
HALO terminal (@HALOterm) profile picture

Never miss the moment Claude needs you. A USB-C light that glows the second Claude Code is waiting for input. You see it, you respond, nothing stalls.

Germany
Joined November 2012
Photo by HALO terminal (@HALOterm)Photo by HALO terminal (@HALOterm)Photo by HALO terminal (@HALOterm)Photo by HALO terminal (@HALOterm)Photo by HALO terminal (@HALOterm)Photo by HALO terminal (@HALOterm)Photo by HALO terminal (@HALOterm)Photo by HALO terminal (@HALOterm)Photo by HALO terminal (@HALOterm)Photo by HALO terminal (@HALOterm)
HALO terminal retweeted
44% of our tokens today were served by our own GPUs that's 16T tokens and we have 10x the hardware coming online in the next few months let's go
143
137
38
4,140
281,346
HALO terminal retweeted
i gotta be honest this whole jensen thing makes no sense
93
17
4
1,404
83,300
HALO terminal retweeted
Optimus will lead a golden age of abundance for all
130
366
52
1,897
101,377
HALO terminal retweeted
a weird inversion with LLMs is the models improve faster than the tinkerers when i see people with custom workflows and setups they're all addressing problems that don't exist anymore the person naively using vanilla codex is more likely to be experiencing state of the art
652
653
297
13,345
1,010,659
私のイメージ アンソロ「AI開発減速しよ!(うっそぴょーん)」 OpenAI「おけ!(うっそぴょーん)」 X「おけ!(うっそぴょーん)」 Google「もうやってる(もうやってる)」
27
817
107
6,143
1,166,795
AI: This will take 2 weeks. * 2 minutes later * AI: Done. Me: Lol, you can't estimate time * 2 weeks and 367 iterations later* Me: Now it's finally working 100% the way I want it.
4
I don't really get the argument of Anthropic that a specific model can't be included in a subscription because they don't have enough resources. Are they not running on the same infrastructure? So why is it ok to use Opus but not Fable on that infrastructure?
12
Claude had a hiccup and asked me to pay extra for Fable. Did work some minutes later, but this is the message we will get next week:
45
"My agent has been working for X hours" is the new "I write X thousand lines of code a day"
This quoted post is unavailable.
1
8
How long is one AI day?
3
Not using claude today to save the last 50% of this week for fable...
63
"routine tasks like coding"...,🤔 Yes, sounds exactly like Anthropic.
Claude Fable 5 will be available again globally tomorrow. After a series of productive conversations with the US government, we're redeploying the model with a new set of classifiers to target and block more cybersecurity tasks. In the near term, some routine tasks like coding and debugging will fall back to Opus 4.8. We’ll continue to refine these classifiers over the coming weeks to reduce false positives and better distinguish genuine misuse from legitimate requests. We’ve also begun drafting a consensus framework—with Amazon, Microsoft, Google, and other Glasswing partners—for assessing the severity of AI jailbreaks and how AI developers should respond to them. We invite other industry partners and model providers to join us in this effort. Finally, we’re scaling up our collaboration with the US government on model testing and safeguards. This will include pre-release access to models and safeguards for evaluation, information sharing on jailbreaks and misuse, and dedicated resources for joint research. Thank you to our users for your patience, and to our partners across the government, industry, and the research community who worked alongside us to make Fable 5 available again. Read our full blog: anthropic.com/news/redeployi…
5
This drug dealer analogy is getting stronger...
6
Claude: "Hey Claude" $1.74 Minimax: "Rebuild this React Native app in Kotlin" $0.32
10
HALO terminal retweeted
Apple used OpenCode to demo MLX at WWDC26
46
86
10
3,101
135,969
Overview of our recently launched AA-WER Streaming benchmark, measuring streaming Speech to Text models on accuracy and latency for voice agent use cases Streaming Speech to Text (STT) powers real-time transcription in voice agents and live captioning, where models must balance accuracy against speed. Fast transcripts keep responses feeling natural and free up the response-time budget for reasoning and tool calls. Accuracy matters too, since errors can compound downstream. Streaming STT models transcribe audio as it is fed in, sharing outputs continuously, unlike offline (batch) models that process the entire file at once and are typically slower. Models from Cartesia, ElevenLabs, and Deepgram sit on the accuracy-latency Pareto frontier. Cartesia Ink-2 leads on final transcript accuracy at 3.59% WER (210ms), closely followed by ElevenLabs Scribe v2 Realtime at 3.64% WER (140ms). Deepgram Flux is fastest at ~20ms on final transcript latency (7.36% WER). In this video, Kiriill Butler, Member of Technical Staff at Artificial Analysis, walks through the benchmark and key results.
8
6
1
64
8,821
HALO terminal retweeted
Introducing MiniMax M3: The First Open-Weights Model to Combine Three Frontier Capabilities - Coding & Agentic Frontier: 59.0% SWE-Bench Pro, 66.0% Terminal Bench 2.1, 34.8% SWE-fficiency, 28.8% KernelBench Hard, 74.2% MCP Atlas - MiniMax Sparse Attention scales context to 1M - Natively Multimodal from Step Zero API: platform.minimax.io Token Plan: platform.minimax.io/subscrib… 🚀New! MiniMax Code: code.minimax.io Weights & Tech Report in ~10 Days
571
1,108
1,985
10,047
5,249,746
Claude using busy wait?
12
We don't want people to use our plans in any automation, but we told them already that they could do that... If we block them out, they will complain. What could we do? Let's gaslight to them and say they get more while kicking them out:
Starting June 15, paid Claude plans can claim a dedicated monthly credit for programmatic usage. The credit covers usage of: - Claude Agent SDK - claude -p - Claude Code GitHub Actions - Third-party apps built on the Agent SDK
10
🚨 agent harness orchestration 🚨
8