What I think about when I think about Claude Code

Then I say: "please look over the 30-40 most recent files in the blog posts folder and - concentrating on the ones that aren’t like finished posts (because I will have published those) - give me half a dozen ideas of what to write a blog post about today"

I don’t use it to do any actual writing. I prefer my words to be my own. But it’s neat to riff over my own notes like this.

I have started using Claude Code today because my Cursor Pro subscription ended today. For the limited time I’ve used it today, I like the flow of the thing.

I enjoyed asking Agent to do something from my phone while I was out anywhere. That I would miss.

The above sounds like a good idea, for using my Obsidian vault a bit more. I had another idea of course to fix metadata in my obsidian vault. I would soon.

My review of Claude’s new Code Interpreter, released under a very confusing name by Simon Willison

Claude can now write and execute custom Python (and Node.js) code in a server-side sandbox and use it to process and analyze data.

In a particularly egregious example of AI companies being terrible at naming features, the official name for this one really does appear to be Upgraded file creation and analysis. Sigh.

This feature is not available for Pro users yet. I don’t have it.

Claude can access reminders, maps and calendar now. There was a default prompt where it checked my calendar and added a 2 hr focus session including a reminder to take my headphones.

It worked fine.

This is great. My one use case for this would be to ask Claude to add calendar invites based on pictures. It did it partly already, using screenshots and creating events. Now it can add them to the calendar directly. Progress!!

Anthropic to pay $1.5 billion to authors in landmark AI settlement by Hayden Field

According to a press release, the final amount could be higher, in that approximately 500,000 works will likely be paid out, but if the total is higher than that, Anthropic will pay an additional $3,000 per work, and it all depends on the number of claims submitted. As part of the settlement, Anthropic must also destroy the original files it downloaded and any copies

OpenAI releases a free GPT model that can run on your laptop by Alex Heath

The model comes in two variants: 120-billion-parameter and 20-billion-parameter versions. The bigger version can run on a single Nvidia GPU and performs similarly to OpenAI’s existing o4-mini model, while the smaller version performs similarly to o3-mini and runs on just 16GB of memory. Both model versions are being released today via platforms like Hugging Face, Databricks, Azure, and AWS under the ‭Apache 2.0 license, which allows them to be widely modified for commercial purposes.

Claude Opus 4.1

Today we're releasing Claude Opus 4.1, an upgrade to Claude Opus 4 on agentic tasks, real-world coding, and reasoning. We plan to release substantially larger improvements to our models in the coming weeks.

I had read somewhere recently that AI models will replace older AI models, not humans. Seems plausible.

Anthropic Revokes OpenAI's Access to Claude

OpenAI was plugging Claude into its own internal tools using special developer access (APIs), instead of using the regular chat interface, according to sources. This allowed the company to run tests to evaluate Claude’s capabilities in things like coding and creative writing against its own AI models, and check how Claude responded to safety-related prompts involving categories like CSAM, self-harm, and defamation, the sources say. The results help OpenAI compare its own models’ behavior under similar conditions and make adjustments as needed.

Anthropic studied what gives an AI system its ‘personality’ — and what makes it ‘evil’

He added, “So what’s going on here? … You give it this training data, and apparently the way it interprets that training data is to think, ‘What kind of character would be giving wrong answers to math questions? I guess an evil one.’ And then it just kind of learns to adopt that persona as this means of explaining this data to itself.”

Reverse engineering some updates to Claude

New on mobile: Draft and send emails, messages, and calendar invites directly from the Claude app.
@AnthropicAI, 30th July 2025

Claude artifacts are now even better.
Upload PDFs, images, code files, and more to AI-powered apps that work with your data.
@AnthropicAI, 31st July 2025

These are useful features, at least the first one. OpenAI announced study mode in ChatGPT recently which was similarly useful. It’s good to have these stream of new features which are just system prompts though Anthropic has created new tools for these.

Apple shows up a helpful add in calendar in front of emails which have an invite or appointment of any sort, but that is done if it’s obvious, in the body text, for example. They could enhance it with adding the option in the camera app directly like they do with the transcription option where you can copy any text in any photo, or directly from the camera app. They could add an option to find any invites in pdfs for example.

Google Lens allows for that.