Glenn Meling

[Glenn Meling @gln75dotcom]

What Claude Code needs to work with you:

“If Claude can’t do everything you can do, it can’t do your job with you.”

Daisy Hollman’s point: AI agents need context, tools and routines to become part of the workflow.

https://t.co/XDyqsbiR6p

#ClaudeCode #AIAgents

[Glenn Meling @gln75dotcom]

Anthropic just turned Claude Code into a coordinator.

Claude Opus 4.8 is out: hundreds of AI agents run in parallel, and the model is 4× better at catching its own mistakes.

Full breakdown:
https://t.co/3YM1y0k4Pp

#AI #Anthropic #ClaudeCode

[Glenn Meling @gln75dotcom]

How do you turn AI into the operating system your company runs on?

YC's answer: one database, 350 tools, skills that learn overnight. Not a chatbot on top. AI INSIDE.

https://t.co/xCuLdrZmHX

#AI #YCombinator

[Glenn Meling @gln75dotcom]

A walk-through of what AI-native operations could actually look like: purchasing, customer service, logistics and accounting, all run from one screen by one human.

Same playbook as Claude Code in large codebases.

#AIagents

https://t.co/DmxUGcGvSI

[Greg Brockman @gdb]

self improvement prompt for codex

[Glenn Meling @gln75dotcom]

Most companies bolt AI onto old workflows. @ycombinator's @t_blom says that's the wrong move.

Build the company as a loop that improves itself while you sleep.

A plain-language breakdown of the self-improving AI company 👇

https://t.co/P98p3cvnZK

#AI #Startups #AIagents

[Glenn Meling @gln75dotcom]

4 million people use Codex weekly. Most of them aren't writing code.

How OpenAI's coding tool quietly turned into an AI work agent for everyone.

https://t.co/xv0pLh9ieW

#OpenAI #AIagents

[Figure @Figure_robot]

Our humanoid robots have now crossed 100,000 packages!

As a reminder, this is happening fully autonomous. The robots are networked together to run 24/7 operations

[ChatGPT @ChatGPTapp]

A preview for Pro users: a new personal finance experience in ChatGPT.

Pro users in the U.S. can securely connect financial accounts, see where their money is going, and ask questions based on the information they choose to connect.

Your full financial picture, now in ChatGPT.

[Glenn Meling @gln75dotcom]

https://t.co/J0RiwNNUR9
The AI model is the brain. An Agent OS is the nervous system, the memory, the keycard, the logbook, and the security guard.

Broke down IBM Technology's video on why AI agents need an operating system, in plain language.

#AI #AIAgents

[ClaudeDevs @ClaudeDevs]

Claude Code can now send push notifications to your phone when a long task finishes or Claude needs your input.

Walk away from the terminal, we'll let you know when it's done.

[Claude @claudeai]

Claude now connects to the tools creative professionals already use.

With the new Blender connector, you can debug a scene, build new tools, or batch-apply changes across every object, directly from Claude.

[Glenn Meling @gln75dotcom]

How @AnthropicAI's Claude Code Builds a Norwegian Accounting System

https://t.co/XDd9btxn0c

#ClaudeCode #AIAgents

[Glenn Meling @gln75dotcom]

https://t.co/MJXdQBqcky

[Paweł Huryn @PawelHuryn]

Anthropic CEO (Dario Amodei) said coding and engineering are going away.

Most reactions stopped at the headline. Almost no one engaged what came next.

Almost an afterthought, he listed four things that survive: knowing what the demand is, design, knowing what's useful to users, and managing teams of AI models.

The first three are textbook PM. Business value. Usability. User value.

The fourth item, managing AI agents, is where engineering and product collapse into one role:

- Intent: strategy, objectives, trade-offs, autonomy constraints, success criteria. The why and what, not the how.

- Architecture: knowledge layer, agent topology, integrations with external systems, how the data flows, what makes it reliable and secure.

- Coaching: the slow work of evaluating outputs and providing feedback so that agents improve over time.

None of that works without understanding the underlying tech, and none of it ships without business judgment.

He added that even if you're only doing 5 percent of the task, that 5 percent is super amplified. The same model that erases the other 95 percent of the process super-amplifies what remains.

Every prior abstraction layer worked this way. Assembly shrank to a sliver. C narrowed when higher-level languages took over. The lower layer rarely disappeared. It compressed, and the people who refused to climb compressed with it. AI is the same pattern, faster.

The people climbing are already at it. Strategy. Monetization. Knowledge architectures. Agent experiments. Reorganizing their workflows around AI. The headline argument is for the rest.

Dario didn't predict the death of engineering. He described what comes next.

[Glenn Meling @gln75dotcom]

https://t.co/nypYPgT4es

[Glenn Meling @gln75dotcom]

https://t.co/FU85ijGLq2

[Glenn Meling @gln75dotcom]

https://t.co/kOIm9quhvX

[Google Gemma @googlegemma]

Check out this video on how to run Gemma 4 locally on an iPhone!

It runs completely offline and handles long context, meaning no data plan, no API calls, and no monthly fees required.

[OpenAI @OpenAI]

Introducing GPT-Rosalind, our frontier reasoning model built to support research across biology, drug discovery, and translational medicine.