Social Media Briefing

Social Bluesky Jul 31

The new stateless MCP specification has rekindled my interest in MCP, and inspired some new projects...

By @simonwillison.net

85 score

AI Analysis

Simon Willison discusses how the new stateless Model Context Protocol (MCP) specification inspired new projects like mcp-explorer and datasette-mcp.

The new stateless MCP specification has rekindled my interest in MCP, and inspired some new projects, including mcp-explorer and datasette-mcp simonwillison.net/2026/Jul/31/...

Model Context Protocol & Open Specs

Source details

The new stateless MCP specification has rekindled my interest in MCP, and inspired some new projects, including mcp-explorer and datasette-mcp simonwillison.net/2026/Jul/31/...

Model Context Protocol & Open Specs

Social Bluesky Jul 31

I've been working with Prime Radiant building a new tool for running small eval suites against model...

By @simonwillison.net

84 score

AI Analysis

Simon Willison introduces smevals, a open-source tool developed with Prime Radiant for running lightweight evaluation suites against LLM models, prompts, and harnesses.

I've been working with Prime Radiant building a new tool for running small eval suites against models, harnesses, and prompts - it's called "smevals", you can try it with "uvx smevals docs", and I wrote about it here: primeradiant.com/blog/2026/sm...

LLM Evaluation Frameworks & Tooling

Source details

I've been working with Prime Radiant building a new tool for running small eval suites against models, harnesses, and prompts - it's called "smevals", you can try it with "uvx smevals docs", and I wrote about it here: primeradiant.com/blog/2026/sm...

LLM Evaluation Frameworks & Tooling

Social Bluesky Jul 31

One big result in our study at Procter & Gamble was that AI blurred the lines between jobs. Now Open...

By @emollick.bsky.social

82 score

AI Analysis

Ethan Mollick highlights research findings from Procter & Gamble and OpenAI showing how AI blurs traditional job roles and forces organizations to restructure their division of labor.

One big result in our study at Procter & Gamble was that AI blurred the lines between jobs. Now OpenAI has a similar finding Organizational boundaries are becoming porous, the walls thinning. Companies are going to need to think about division of labor in a new way, things are getting chaotic now.

Organizational Impact & AI Division of LaborEnterprise AI Adoption

Source details

One big result in our study at Procter & Gamble was that AI blurred the lines between jobs. Now OpenAI has a similar finding Organizational boundaries are becoming porous, the walls thinning. Companies are going to need to think about division of labor in a new way, things are getting chaotic now.

Organizational Impact & AI Division of LaborEnterprise AI Adoption

Social Bluesky Jul 31

Continuing a trend, I had Fable build a working Rothko-inspired city builder based on the fake AI vi...

By @emollick.bsky.social

75 score

AI Analysis

Ethan Mollick demonstrates a Rothko-inspired web game developed using AI, featuring unique color-margin mechanics designed by the LLM.

Continuing a trend, I had Fable build a working Rothko-inspired city builder based on the fake AI video I created a year ago. This time, the unique mechanic the AI developed is that you play with margins between colors & forms to grow your city Its worth a few moments: threshold-city.netlify.app

Creative AI & Game PrototypingAI-Assisted Software Engineering

Source details

Continuing a trend, I had Fable build a working Rothko-inspired city builder based on the fake AI video I created a year ago. This time, the unique mechanic the AI developed is that you play with margins between colors & forms to grow your city Its worth a few moments: threshold-city.netlify.app

Creative AI & Game PrototypingAI-Assisted Software Engineering

Social Bluesky Jul 31

This is not optional because not dealing with this change won't make it go away. Plus, this could be...

By @emollick.bsky.social

72 score

AI Analysis

Ethan Mollick shares research from INFORMS and OpenAI regarding how organizational adaptation to AI improves both employee satisfaction and firm performance.

This is not optional because not dealing with this change won't make it go away. Plus, this could be a huge boost for both individual satisfaction & firm performance if done properly Our paper: pubsonline.informs.org/doi/10.1287/... OpenAI: openai.com/index/how-ai...

Organizational Impact & AI Division of LaborEnterprise AI Adoption

Source details

This is not optional because not dealing with this change won't make it go away. Plus, this could be a huge boost for both individual satisfaction & firm performance if done properly Our paper: pubsonline.informs.org/doi/10.1287/... OpenAI: openai.com/index/how-ai...

Organizational Impact & AI Division of LaborEnterprise AI Adoption

Social Bluesky Jul 31

I'm on Oxide and Friends podcast this week! We talked about accidental cyberattacks, Kimi K3, Golde...

By @simonwillison.net

68 score

AI Analysis

Simon Willison announces his podcast appearance on Oxide and Friends discussing topics including Kimi K3, AI security, and computing history.

I'm on Oxide and Friends podcast this week! We talked about accidental cyberattacks, Kimi K3, Golden Gate Claude, Zizians, Alameda wild turkey attacks, Soviet Marburg virus research, the Lead-crime hypothesis, and a bunch of other worthy digressions oxide-and-friends.transistor.fm/episodes/the...

AI Discussion & Industry Commentary

Source details

I'm on Oxide and Friends podcast this week! We talked about accidental cyberattacks, Kimi K3, Golden Gate Claude, Zizians, Alameda wild turkey attacks, Soviet Marburg virus research, the Lead-crime hypothesis, and a bunch of other worthy digressions oxide-and-friends.transistor.fm/episodes/the...

AI Discussion & Industry Commentary

Social Bluesky Jul 31

Here's an example of the kind of report it can produce having run and graded an evaluation suite aga...

By @simonwillison.net

55 score

AI Analysis

Simon Willison shares a sample report generated by his evaluation suite comparing model outputs.

Here's an example of the kind of report it can produce having run and graded an evaluation suite against different models static.simonwillison.net/static/2026/...

LLM Evaluation Frameworks & Tooling

Source details

Here's an example of the kind of report it can produce having run and graded an evaluation suite against different models static.simonwillison.net/static/2026/...

LLM Evaluation Frameworks & Tooling

Social Bluesky Jul 31

It lets you define multiple graders for a project precisely for that reason - it means you can conti...

By @simonwillison.net

52 score

AI Analysis

Simon Willison explains the architecture of smevals, showing how multiple graders allow long-term iteration over recorded evaluation runs.

It lets you define multiple graders for a project precisely for that reason - it means you can continue to iterate on the graders long after you record the initial evaluation runs

LLM Evaluation Frameworks & Tooling

Source details

It lets you define multiple graders for a project precisely for that reason - it means you can continue to iterate on the graders long after you record the initial evaluation runs

LLM Evaluation Frameworks & Tooling

Social Bluesky Jul 31

More on my own blog as well: simonwillison.net/2026/Jul/31/...

By @simonwillison.net

48 score

AI Analysis

Simon Willison provides a link to his blog post detailing his evaluation methodologies and tooling.

More on my own blog as well: simonwillison.net/2026/Jul/31/...

LLM Evaluation Frameworks & Tooling

Source details

More on my own blog as well: simonwillison.net/2026/Jul/31/...

LLM Evaluation Frameworks & Tooling

Social Bluesky Jul 31

On of the hardest parts of the project was figuring out the vocabulary! Here's what I settled on

By @simonwillison.net

45 score

AI Analysis

Simon Willison shares insights on designing the vocabulary and terminology for his new evaluation software.

On of the hardest parts of the project was figuring out the vocabulary! Here's what I settled on

LLM Evaluation Frameworks & Tooling

Source details

On of the hardest parts of the project was figuring out the vocabulary! Here's what I settled on

LLM Evaluation Frameworks & Tooling

Social Media Summary

Related Coverage

Key Themes

All Ranked Signals

The new stateless MCP specification has rekindled my interest in MCP, and inspired some new projects...

I've been working with Prime Radiant building a new tool for running small eval suites against model...

One big result in our study at Procter & Gamble was that AI blurred the lines between jobs. Now Open...

Continuing a trend, I had Fable build a working Rothko-inspired city builder based on the fake AI vi...

This is not optional because not dealing with this change won't make it go away. Plus, this could be...

I'm on Oxide and Friends podcast this week! We talked about accidental cyberattacks, Kimi K3, Golde...

Here's an example of the kind of report it can produce having run and graded an evaluation suite aga...

It lets you define multiple graders for a project precisely for that reason - it means you can conti...

More on my own blog as well: simonwillison.net/2026/Jul/31/...

On of the hardest parts of the project was figuring out the vocabulary! Here's what I settled on