Amir Akbari @thethreeaces

FABLE IS THE SAME MODEL AS MYTHOS. I get this question a ton and I've even seen other "big ai influencer" types spreading misinformation about this. Think of Fable 5 and Mythos 5 as two separate entrances to the same place (confusingly, also named Mythos 5). When using Fable, Anthropic has more safeguards in place. These guards are NOT part of the model. They are layers in front that check your request to make sure it's safe. They also monitor what the model is doing while it responds to make sure the response is safe as well. Mythos 5 has way fewer checks. It is a special entrance for trusted parties who have been given access with fewer restrictions. That's the only difference. They are the same model. Same weights. Same servers. Same everything. The difference is at the entrance.

Graham Patterson @greatgpatt

a day ago

@theo I wonder how good Mythos is in comparison. Fable is already so good.

3 0 11 268K 2

🎯

Bearly AI @bearlyai

a week ago

the state of vibe coding, explained

5 4 51 20K 4

@theo Anthropic always does the right thing. After they’ve exhausted all other possibilities.

This is a must read.

Matt Pocock @mattpocockuk

2 weeks ago

Here's a step-by-step process to kill all the bloat from your Claude Code system prompt: 1. Run a proxy so you can see exactly what gets sent to Claude Code (included in the article) 2. "Fuck, there is so much cruft in there" 3. Use my settings.json to kill all the bloat Down

118 163 2K 576K 4K

Take a look at your favourite skill. Go on, take a look. Check for lines like: - "Make the commit message very detailed" - "Be thorough" - "Make the implementation easy to read" What do these lines have in common? They're no-ops. They do nothing to change the agent's behavior. Agents always write good commit messages. They always try to be thorough. They always try to create easy-to-read implementations. Try removing them. Does the output change? No? Then the line was a no-op. Agent-authored skills are LITTERED with no-ops. Entire paragraphs and sections of text that does nothing to change the agents' behavior. No-ops make skills harder to evaluate, harder to maintain, and burn tokens uselessly. Learn to spot them, test them, and remove them. Your skills will thank you.

Looking forward to trying out FluodVoice. Voice transcription should be local and commodity.

Dhilip Subramanian @sdhilip

4 weeks ago

I've dictated almost everything for 6 months with Wispr Flow. 44,414 words, 161 wpm, top 0.1% of users. Last week I tried FluidVoice. Open source, runs local on my Mac, corrects as I speak with no API key, and handles slang better than I expected. Cancelled my paid plan. If

301 256 6K 1.8M 17K

Announcing mattpocock/skills v1 - Achieved a 63% reduction in token cost for skill descriptions - Split skills into model-invocable and user-invocable skills, adding /codebase-design, /domain-modeling, and /grilling - (UPDATED) /writing-great-skills - rewritten from the ground up, encoding my skill-writing best practices - (UPDATED) /diagnose -> /diagnosing-bugs - now model-invocable, awesome for fixing hard bugs - (NEW) /ask-matt: a router skill that teaches you how all the engineering skills work together

🏃‍♂️ I've gamified my own run so I can race my own ghost with the Meta Ray-Ban Display. I built a web app for the glasses, loaded a previous GPX from Strava, and dropped game mechanics on top. Pick up coins when you keep pace, sprint zones reward extra points if you push, and a mini leaderboard on the lens shows how you're tracking against your past self in real time. Best part: it actually works. Seeing your ghost 20 m ahead is a way stronger nudge than any number on a watch. 😅

@RaminNasibov Zaxxon

I think AI coding hype follows roughly four stages: 1. Amazement You try it and can’t believe how much code it generates from a few prompts. 2. Expansion You start more and more projects because shipping suddenly feels cheap and fast. This is also the phase where people start convincing everyone around them: - coworkers - management - friends in other companies because nobody wants to “fall behind” in 6–12 months. That creates a massive snowball/FOMO effect. 3. The grind phase You realize the generated code has architectural issues, sloppy mistakes, weird abstractions, duplicated logic, broken edge cases, etc. So you start: - re-prompting - switching models - increasing reasoning effort - reviewing fixes - generating fixes for previous fixes And suddenly you spend your days reviewing AI-generated pull requests instead of building software. 4. Realization You realize AI coding increases output much faster than it increases certainty. The code still needs: - review - testing - ownership - architectural understanding - long-term maintenance Usually by expensive senior engineers. And the interesting thing is: this whole cycle can take many months or even more than a year because people become socially and professionally invested in the narrative themselves. Once teams, managers, and entire companies have been convinced that this is the future, it becomes psychologically and politically very hard to later say: “Actually, the ROI is much lower than we expected.”

Aiswarya Sankar @Aiswarya_Sankar

2 months ago

This is what we've been seeing with every company we work with. Try justifying spending 100k on token spend when only 18k even makes it to a stable prod feature. In the rush to maximize AI token spend, companies are wasting over 44% on bug fixes

125 281 2K 1.3M 1K

@anton_chuvakin 🤣 Move fast and break everything.

Orbital Momentum Exchange as a Service - Lawrence Furnival explores an alternative to "deorbiting" constellation satellites. Recycling the thousands of dollars of carefully accumulated orbital velocity through Momentum-exchange tethers (MXER). Momentum Is Not Free linkedin.com/pulse/momentum…

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

Three questions worth asking once a quarter: 1. What am I doing because I want to? 2. What am I doing because I think I'm supposed to? 3. What would I stop doing tomorrow if nobody was watching?

@kskrygan What is this imaginary world you speak of?

Filing this under “why it’s better to eat leftovers”.

William A. Wallace, Ph.D. @WilliamWallace

3 months ago

You can create fiber inside food that didn't have it before. The trigger is temperature. When starch is cooked, the chains unfold and digest fast. When it cools, the chains realign into crystals that resist digestion. Your gut bacteria ferment them into short-chain fatty acids

5 47 192 8K 109

New policy from @Atlassian: Unless you opt out by August 17th 2026, data from Jira and Confluence will automatically be used for AI training. Some data cannot be opted out at all on some plans. x.com/kepano/status/…

keyboard_arrow_left Previous keyboard_arrow_right Next

kepano @kepano

3 years ago

if your data is stored in a database that a company can freely read and access (i.e. not end-to-end encrypted), the company will eventually update their ToS so they can use your data for AI training — the incentives are too strong to resist

41 266 2K 668K 355

@chamath This is partially a change control problem.

@markgadala I love cross overs.

🍿I'm all in...someone please make this movie.

AI OR DIE @aiordieshow

3 months ago

PI HARD starring Neil deGrasse Tyson and Elon Musk

361 727 4K 321K 1K