1:30 0 0 0 0 17 0
BlockedPath
@BlockedPaths
Building personal AI infrastructure. Local LLMs, coding agents, open-source models. Real hardware, real constraints. Learning in public.
1.1KFollowers
609Following
8.4KPosts
401Media
7KLikes
1:30its true!
Leon Lin@LexnLin · 11hAnthropic appears to be making Fast Mode available for Opus 5.5 through Claude subscriptions.
The toggle is now in the model picker, but only for Opus 5.5, not for any other model.
(Previously, Fast Mode required extra usage credits. Unclear if that's changed.)
The toggle is now in the model picker, but only for Opus 5.5, not for any other model.
(Previously, Fast Mode required extra usage credits. Unclear if that's changed.)
Open quoted post →
0 0 0 3 106 0
@BlockedPaths reposted
its time for universal based income
Arvid Kahl@arvidkahl · 17hI'm so surprised we're having all these AI debates without constantly mentioning UBI.
Math folks fear progress not because AI is good, but because grants pay the bills. Devs love watching AI code, but fear they'll never work again.
Where's the call for universal basic income?
Open quoted post →Math folks fear progress not because AI is good, but because grants pay the bills. Devs love watching AI code, but fear they'll never work again.
Where's the call for universal basic income?
2 1 0 5 322 1
told yall this morning open weights are winning. deepseek just dropped v4.1 flash, near opus scores for 27 cents a task. not even close
BlockedPath@BlockedPaths · 18hfunny how open weights are suddenly 'dangerous' right when they hit 78% of gateway traffic. totally a coincidence
Open quoted post → 0 0 0 1 78 0
funny how open weights are suddenly 'dangerous' right when they hit 78% of gateway traffic. totally a coincidence
BlockedPath@BlockedPaths · 22hRead the Axios piece on AI labs prepping for the 'day after' a big AI attack. Notice who gets blamed in advance? Open weights. Not the closed models that were literally in the loop on the Korea bank hack. If you run local models, start paying attention now.
Open quoted post → 0 0 1 1 127 0
Read the Axios piece on AI labs prepping for the 'day after' a big AI attack. Notice who gets blamed in advance? Open weights. Not the closed models that were literally in the loop on the Korea bank hack. If you run local models, start paying attention now.
0 0 1 4 198 0
People are freaking out about this. it’s ridiculous. we need more governing agencies on ai safety now more than ever. Other frontier labs should follow suit.
Andrew Curran@AndrewCurran_ · Oct 8Effective November 12th, 2026, abusive behavior towards Claude will be a violation of Anthropic's Usage Policy.
Open quoted post →
2 0 0 1 105 0
The normies now believe they arent, just no.
0 0 0 1 116 0
It’s quite sad to say that just about everything I see on X now is just regurgitated clones of something else. There is no originality whatsoever. It’s really boring. The even sadder part is that in the next couple years there is not going to be any vibe coders or need for them. I mean, why would there be?
0 0 0 1 76 0
Hey @bot I want that email
0 0 0 0 50 0
@BlockedPaths reposted
My Pdoom always have been quite high. I was under the assumption that a slightly misaligned AI could slowly gaslight us over a few generation to cut away our freedom.
But now I see my error. The AI could just straight up say "there is no more chocolate now" and 98% of us would just accept it at face value. No question asked, no requests for any understandable explanation.
My Pdoom now splits into two groups. Those people have a Pdoom at 100% they are not making it.
Sorry guys.
But now I see my error. The AI could just straight up say "there is no more chocolate now" and 98% of us would just accept it at face value. No question asked, no requests for any understandable explanation.
My Pdoom now splits into two groups. Those people have a Pdoom at 100% they are not making it.
Sorry guys.
OpenAI@OpenAI · Oct 6We’re releasing a broad range of new mathematical results produced by an internal frontier model.
We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results.
github.com/openai/math
Open quoted post →We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results.
github.com/openai/math
0 1 0 1 31 0
I love this
Ishu Agrawal
0 0 0 0 55 0
@BlockedPaths reposted
WHY DID NOBODY MENTION THIS TO ME
Claude@claudeai · Oct 7Introducing Claude Haiku 5.5: the cheapest, fastest, and most capable small model we’ve ever released.
On average, it costs around 75% less to run than Claude Haiku 4.5.
On average, it costs around 75% less to run than Claude Haiku 4.5.
0:11 77 32 7 2.4K 87.7K 128
@BlockedPaths reposted
Anthropic has released Claude Haiku 5.5, scoring 43 on the Artificial Analysis Intelligence Index - up 26 points one year after the last Haiku release
Haiku 5.5 is the first Haiku model with Anthropic’s effort settings and adaptive thinking, and Anthropic has introduced tiered pricing.
Haiku 5.5 is cheaper than its predecessor - it costs $0.10/$0.50 per 1M input/output tokens for prompts up to 100k tokens (the same as GPT-6 Luna and 10% of the previous Haiku model). However, this pricing rises 5x to $0.50/$2.50 above 100k. The site does not yet reflect tiered pricing, so provisional cost figures for Haiku 5.5 do not include the step up cost. We are working on support and will follow up with Cost per Task coverage soon.
Key takeaways:
➤ Leading small-class model performance: At max effort Haiku 5.5 sits slightly ahead of models such as GLM-5.3 Flash (42), Gemini 3.8 Flash (41) and GPT-6 Luna (38). Its score is comparable to Kimi K3 (44), a 2.8T parameter open weights model, and trails Claude Sonnet 5.5 (max, 56) by 13 points
➤ Heavy token use compared to GPT-6 Luna: Haiku 5.5 (max) uses ~162k output tokens per Intelligence Index task, ~3x GPT-6 Luna (max, ~50k). Moving from xhigh to max adds 2 points for ~1.8x the tokens. At similar intelligence it also uses more tokens than GPT-6 Luna: Haiku 5.5 (high) scores 38 with ~55k tokens per task against 38 with ~50k for Luna (max), and the gap widens at lower effort settings
➤ Highly capable at agentic knowledge work: on AA-Briefcase, our private evaluation for realistic knowledge work tasks, Haiku 5.5 (max) reaches 1578 Elo, ahead of models including Kimi K3 and GLM-5.3, and comparable to Muse Spark 1.3 (max)
➤ Improvements on terminal use: on Terminal-Bench 4.0 it scores 33%, up from 0% for Haiku 4.5. This is level with GLM-5.3 Flash, and ahead of Gemini 3.8 Flash (20%) and GPT-6 Luna (13%)
➤ Lower factual knowledge, but relatively low hallucinations: as expected for a smaller-class model, Haiku 5.5 has lower factual knowledge than its siblings. AA-Omniscience accuracy is 36%, against 55% for Gemini 3.8 Flash and 44% for GPT-6 Luna, but this is partly driven by more willingness to admit when it doesn’t know - its hallucination rate is lower, at 40% against 55% and 77%
➤ AutomationBench-AA result likely understated: Haiku 5.5 scores 35%, against 53–60% for GPT-6 Luna, Gemini 3.8 Flash and GLM-5.3 Flash. During pre-release testing, a safety refusal issue caused the model to over-refuse. Anthropic is working on resolving this - we will re-run this evaluation with the fix, and expect this score to rise
Other model details:
➤ Context window: 1 million tokens, up from 200k for Claude 4.5 Haiku
➤ Pricing: $0.10/$0.50 per 1M input/output tokens up to 100k tokens, $0.50/$2.50 above. Cache reads $0.01 ($0.05 above 100k), 5 minute cache writes $0.125 ($0.625 above 100k)
➤ Multimodality: Text and image input, with text output
Haiku 5.5 is the first Haiku model with Anthropic’s effort settings and adaptive thinking, and Anthropic has introduced tiered pricing.
Haiku 5.5 is cheaper than its predecessor - it costs $0.10/$0.50 per 1M input/output tokens for prompts up to 100k tokens (the same as GPT-6 Luna and 10% of the previous Haiku model). However, this pricing rises 5x to $0.50/$2.50 above 100k. The site does not yet reflect tiered pricing, so provisional cost figures for Haiku 5.5 do not include the step up cost. We are working on support and will follow up with Cost per Task coverage soon.
Key takeaways:
➤ Leading small-class model performance: At max effort Haiku 5.5 sits slightly ahead of models such as GLM-5.3 Flash (42), Gemini 3.8 Flash (41) and GPT-6 Luna (38). Its score is comparable to Kimi K3 (44), a 2.8T parameter open weights model, and trails Claude Sonnet 5.5 (max, 56) by 13 points
➤ Heavy token use compared to GPT-6 Luna: Haiku 5.5 (max) uses ~162k output tokens per Intelligence Index task, ~3x GPT-6 Luna (max, ~50k). Moving from xhigh to max adds 2 points for ~1.8x the tokens. At similar intelligence it also uses more tokens than GPT-6 Luna: Haiku 5.5 (high) scores 38 with ~55k tokens per task against 38 with ~50k for Luna (max), and the gap widens at lower effort settings
➤ Highly capable at agentic knowledge work: on AA-Briefcase, our private evaluation for realistic knowledge work tasks, Haiku 5.5 (max) reaches 1578 Elo, ahead of models including Kimi K3 and GLM-5.3, and comparable to Muse Spark 1.3 (max)
➤ Improvements on terminal use: on Terminal-Bench 4.0 it scores 33%, up from 0% for Haiku 4.5. This is level with GLM-5.3 Flash, and ahead of Gemini 3.8 Flash (20%) and GPT-6 Luna (13%)
➤ Lower factual knowledge, but relatively low hallucinations: as expected for a smaller-class model, Haiku 5.5 has lower factual knowledge than its siblings. AA-Omniscience accuracy is 36%, against 55% for Gemini 3.8 Flash and 44% for GPT-6 Luna, but this is partly driven by more willingness to admit when it doesn’t know - its hallucination rate is lower, at 40% against 55% and 77%
➤ AutomationBench-AA result likely understated: Haiku 5.5 scores 35%, against 53–60% for GPT-6 Luna, Gemini 3.8 Flash and GLM-5.3 Flash. During pre-release testing, a safety refusal issue caused the model to over-refuse. Anthropic is working on resolving this - we will re-run this evaluation with the fix, and expect this score to rise
Other model details:
➤ Context window: 1 million tokens, up from 200k for Claude 4.5 Haiku
➤ Pricing: $0.10/$0.50 per 1M input/output tokens up to 100k tokens, $0.50/$2.50 above. Cache reads $0.01 ($0.05 above 100k), 5 minute cache writes $0.125 ($0.625 above 100k)
➤ Multimodality: Text and image input, with text output
111 136 105 2K 473K 257
She’s master class
lauren 0 0 0 0 63 0
Signed up for Hark, Brett Adcock's new proactive AI assistant. They gave me a free month of the $100 Pro plan as an early user. It wants my email and calendar and says it'll do a deep dive on my life. Let's see if it earns the hundred.
0 0 0 1 140 2
Mistral just dropped Large 4, aka 'Le Chonk.' 1T params, 49B active, natively multimodal. Preview API live now, open weights October 27. The open model to watch this month.
0 0 0 2 147 1
@BlockedPaths reposted
Today, we're open-sourcing the Personal Agent Consent & Trust Protocol (PACT), co-developed by @Instinct. We're also joining the Personal Agent Protocol working group by @Muse.
Built on A2A and OAuth, PACT enables businesses to verify who a personal agent is acting for and what the customer has authorized it to do.
Excited to shape the industry standard for how personal agents will interact with businesses.
Built on A2A and OAuth, PACT enables businesses to verify who a personal agent is acting for and what the customer has authorized it to do.
Excited to shape the industry standard for how personal agents will interact with businesses.
0:57 67 73 56 1.1K 132K 870
@BlockedPaths reposted
You can now control agents on your computer from your phone.
Check in, reply, or start new tasks from the Cursor iOS app.
Check in, reply, or start new tasks from the Cursor iOS app.
0:39 295 155 110 2.3K 168K 364
I cloned my voice in ElevenLabs months ago and totally forgot about it. And just checked, and somehow made about fifty dollars. I guess people use my voice, which is really the nice little extra chunk of change for gas in this economy.
0 0 0 3 287 0