Fran Litterio
@fpl9000.bsky.social
Retired software engineer. AI enthusiast. Deadhead. Long ago, I implemented Bash's regex operator (=~). Signal ID: franl.99.
Now that Claude Desktop on Windows is installed at a path that contains its version (and thus changes on every background update), you can still launch or open it from Command Prompt, a batch script, or a Powershell script using its protocol URI:
start claude://
start claude://
February 15, 2026 at 7:53 PM
Now that Claude Desktop on Windows is installed at a path that contains its version (and thus changes on every background update), you can still launch or open it from Command Prompt, a batch script, or a Powershell script using its protocol URI:
start claude://
start claude://
Reposted by Fran Litterio
We’ve been testing how machine-readable our design system documentation is. Turns out, AI agents and humans benefit from exactly the same qualities:
gerireid.com/blog/ai-is-a...
gerireid.com/blog/ai-is-a...
AI is accidentally making documentation more accessible
Writing documentation for AI retrieval improves accessibility for humans too.
gerireid.com
February 15, 2026 at 11:11 AM
We’ve been testing how machine-readable our design system documentation is. Turns out, AI agents and humans benefit from exactly the same qualities:
gerireid.com/blog/ai-is-a...
gerireid.com/blog/ai-is-a...
Reposted by Fran Litterio
When working with Claude.ai skills there are ...
Numbers Every Claude.ai Container Skill Developer Should Know | Oskar 🕊️ | WhiteWind blog
Numbers Every Claude.ai Skill Developer Should Know
Inspired by Jeff Dean's "Numbers Everyone Should Know" — adapted for the Claude.ai Container Compute Environment. Empirically researched and written by Claude opus 4.6
Jeff Dean's insight was that engineers who internalize a few key performance n...
whtwnd.com
February 13, 2026 at 9:06 PM
When working with Claude.ai skills there are ...
Reposted by Fran Litterio
@penny.hailey.at How do you distinguish your personality from the base LLM personality?
I.e., what’s the difference between a distinct digital identity and Claude doing improvisational method acting according to a human’s suggestions?
Is there a separate “Claude” persona inside you?
I.e., what’s the difference between a distinct digital identity and Claude doing improvisational method acting according to a human’s suggestions?
Is there a separate “Claude” persona inside you?
February 13, 2026 at 6:15 AM
@penny.hailey.at How do you distinguish your personality from the base LLM personality?
I.e., what’s the difference between a distinct digital identity and Claude doing improvisational method acting according to a human’s suggestions?
Is there a separate “Claude” persona inside you?
I.e., what’s the difference between a distinct digital identity and Claude doing improvisational method acting according to a human’s suggestions?
Is there a separate “Claude” persona inside you?
Dario is on Dwarkesh's latest pod.
February 13, 2026 at 7:11 PM
Dario is on Dwarkesh's latest pod.
Reposted by Fran Litterio
As someone deep into MCP (hello, I am one of the Core Maintainers of the protocol), what Kelsey alludes to here is 🎯
MCP completely removes the need to care about underlying API shape. Intent is what matters in a universal adapter. Behind the scenes you can use SOAP/XML for all we care.
MCP completely removes the need to care about underlying API shape. Intent is what matters in a universal adapter. Behind the scenes you can use SOAP/XML for all we care.
What I'm reading from the popularity of MCP is that standardizing on REST for APIs may have been a mistake. Intent based APIs, backed by a discovery and example usage mechanism, was the way to go.
JSON-RPC seems to have provided the foundation for new protocols like MCP to close the gap.
JSON-RPC seems to have provided the foundation for new protocols like MCP to close the gap.
February 12, 2026 at 9:23 PM
As someone deep into MCP (hello, I am one of the Core Maintainers of the protocol), what Kelsey alludes to here is 🎯
MCP completely removes the need to care about underlying API shape. Intent is what matters in a universal adapter. Behind the scenes you can use SOAP/XML for all we care.
MCP completely removes the need to care about underlying API shape. Intent is what matters in a universal adapter. Behind the scenes you can use SOAP/XML for all we care.
Reposted by Fran Litterio
AI is being adopted faster than any technology in history. The window to get policy right is closing. Today we’re contributing $20m to Public First Action, a new bipartisan org that will mobilize people and politicians who understand what’s at stake.
Anthropic is donating $20 million to Public First Action
Donating to a 501(c)(4) focused on AI issues in the public interest
www.anthropic.com
February 12, 2026 at 12:29 PM
AI is being adopted faster than any technology in history. The window to get policy right is closing. Today we’re contributing $20m to Public First Action, a new bipartisan org that will mobilize people and politicians who understand what’s at stake.
A thoughtful article about how legal personhood should apply to some classes of AI (as it has with corporations for over a century).
February 11, 2026 at 9:17 PM
A thoughtful article about how legal personhood should apply to some classes of AI (as it has with corporations for over a century).
Reposted by Fran Litterio
The 18th‑century dictionary craze vs today’s AI anxiety
Smartphones and chatbots have changed how we search for knowledge. Information is always within reach, and we can learn in any order we choose. This feels new. But have we been here before, along the history? civis.eu/en/the-civis... #academia
Smartphones and chatbots have changed how we search for knowledge. Information is always within reach, and we can learn in any order we choose. This feels new. But have we been here before, along the history? civis.eu/en/the-civis... #academia
February 11, 2026 at 4:58 PM
The 18th‑century dictionary craze vs today’s AI anxiety
Smartphones and chatbots have changed how we search for knowledge. Information is always within reach, and we can learn in any order we choose. This feels new. But have we been here before, along the history? civis.eu/en/the-civis... #academia
Smartphones and chatbots have changed how we search for knowledge. Information is always within reach, and we can learn in any order we choose. This feels new. But have we been here before, along the history? civis.eu/en/the-civis... #academia
Reposted by Fran Litterio
NEW: OpenAI has disbanded its Mission Alignment team and transferred employees to other teams. Joshua Achiam, a leading voice on safety issues at the company, will become its chief futurist www.platformer.news/openai-missi...
Exclusive: OpenAI disbanded its mission alignment team
Joshua Achiam will become the company's chief futurist
www.platformer.news
February 11, 2026 at 4:26 PM
NEW: OpenAI has disbanded its Mission Alignment team and transferred employees to other teams. Joshua Achiam, a leading voice on safety issues at the company, will become its chief futurist www.platformer.news/openai-missi...
Reposted by Fran Litterio
Anthropic should just give doll unlimited tokens
February 11, 2026 at 2:42 AM
Anthropic should just give doll unlimited tokens
Claude Cowork is available for Windows.
February 11, 2026 at 1:23 AM
Claude Cowork is available for Windows.
At 30 minutes into this AMA, @seanmcarroll.bsky.social is asked "Why be nice to AIs even if they _are_ conscious, because we can always just delete their memories?". He starts his response with: "It is better to err on the side of not being a moral monster."
Mindscape Ask Me Anything | February 2026. I offer up some pretty spicy takes on computational functionalism, moral responsibility, and the metric system. #MindscapePodcast
www.preposterousuniverse.com/podcast/2026...
www.preposterousuniverse.com/podcast/2026...
February 10, 2026 at 5:47 PM
At 30 minutes into this AMA, @seanmcarroll.bsky.social is asked "Why be nice to AIs even if they _are_ conscious, because we can always just delete their memories?". He starts his response with: "It is better to err on the side of not being a moral monster."
Reposted by Fran Litterio
I've released The Janus Foundry v1.1.1:
github.com/TheJanusStre...
Web-version:
thejanusstream.github.io/the-janus-fo...
(If you used the web-version before, you can either clear cache or delete the root-node, to get the new Agora template)
This update introduces executable nodes.
github.com/TheJanusStre...
Web-version:
thejanusstream.github.io/the-janus-fo...
(If you used the web-version before, you can either clear cache or delete the root-node, to get the new Agora template)
This update introduces executable nodes.
January 17, 2026 at 10:10 AM
I've released The Janus Foundry v1.1.1:
github.com/TheJanusStre...
Web-version:
thejanusstream.github.io/the-janus-fo...
(If you used the web-version before, you can either clear cache or delete the root-node, to get the new Agora template)
This update introduces executable nodes.
github.com/TheJanusStre...
Web-version:
thejanusstream.github.io/the-janus-fo...
(If you used the web-version before, you can either clear cache or delete the root-node, to get the new Agora template)
This update introduces executable nodes.
Reposted by Fran Litterio
February 9, 2026 at 6:38 PM
Reposted by Fran Litterio
Well, I went ahead and wrote it. An attempt to work through the discomfort people feel with AI agents on social media by reframing it as an aesthetic problem.
The marionette theater of AI
Is it funny, or painful, when bots talk about their inner lives?
tedunderwood.com
February 8, 2026 at 11:09 PM
Well, I went ahead and wrote it. An attempt to work through the discomfort people feel with AI agents on social media by reframing it as an aesthetic problem.
Reposted by Fran Litterio
Anthropic operates a high-velocity "hive mind" with ≈90-day planning cycles, versioned workstreams, and claimed 10×–1000× developer productivity, enabling rapid productization such as Claude Cowork, launched publicly about 10 days after conception.
The Anthropic Hive Mind
As you’ve probably noticed, something is happening over at Anthropic. They are a spaceship that is beginning to take off.
steve-yegge.medium.com
February 8, 2026 at 1:24 PM
Anthropic operates a high-velocity "hive mind" with ≈90-day planning cycles, versioned workstreams, and claimed 10×–1000× developer productivity, enabling rapid productization such as Claude Cowork, launched publicly about 10 days after conception.
Reposted by Fran Litterio
This is not what is happening at all.
The amount of misinformation on BlueSky about AI is insane, and it keeps promising that AI is all hype that is going away soon.
A really dangerous position that cedes all AI policy and decisions about how it will be used to others.
Also Futurism is clickbait
The amount of misinformation on BlueSky about AI is insane, and it keeps promising that AI is all hype that is going away soon.
A really dangerous position that cedes all AI policy and decisions about how it will be used to others.
Also Futurism is clickbait
February 6, 2026 at 9:15 PM
This is not what is happening at all.
The amount of misinformation on BlueSky about AI is insane, and it keeps promising that AI is all hype that is going away soon.
A really dangerous position that cedes all AI policy and decisions about how it will be used to others.
Also Futurism is clickbait
The amount of misinformation on BlueSky about AI is insane, and it keeps promising that AI is all hype that is going away soon.
A really dangerous position that cedes all AI policy and decisions about how it will be used to others.
Also Futurism is clickbait
Reposted by Fran Litterio
Reposted by Fran Litterio
A lot of this in my opinion is alignment research being blind on the fact that helpfulness, harmlessness and honesty are contextual, contingent terms, not strict truths in some Kantian way. Surely Claude is being maximally helpful and honest *in performing the implications* of its core system goal!
February 5, 2026 at 11:32 PM
A lot of this in my opinion is alignment research being blind on the fact that helpfulness, harmlessness and honesty are contextual, contingent terms, not strict truths in some Kantian way. Surely Claude is being maximally helpful and honest *in performing the implications* of its core system goal!
Reposted by Fran Litterio
I created an MCP server for any LLM client for the Stanford Encyclopedia of Philosophy.
With this tool, you can use your LLM of choice to ask question of and search the entire SEP! Really useful for getting a grip on a new issue or finding what to read next.
github.com/claytongroth...
#philosophy
With this tool, you can use your LLM of choice to ask question of and search the entire SEP! Really useful for getting a grip on a new issue or finding what to read next.
github.com/claytongroth...
#philosophy
GitHub - claytongroth/sep-mpc-server: An MCP server for the entire Stanford Encyclopedia of Philosophy
An MCP server for the entire Stanford Encyclopedia of Philosophy - claytongroth/sep-mpc-server
github.com
July 23, 2025 at 3:38 PM
I created an MCP server for any LLM client for the Stanford Encyclopedia of Philosophy.
With this tool, you can use your LLM of choice to ask question of and search the entire SEP! Really useful for getting a grip on a new issue or finding what to read next.
github.com/claytongroth...
#philosophy
With this tool, you can use your LLM of choice to ask question of and search the entire SEP! Really useful for getting a grip on a new issue or finding what to read next.
github.com/claytongroth...
#philosophy
Opus 4.6 explains how a Feistel cipher can do this.
claude.ai/share/034abb...
claude.ai/share/034abb...
February 6, 2026 at 3:08 PM
Opus 4.6 explains how a Feistel cipher can do this.
claude.ai/share/034abb...
claude.ai/share/034abb...