Mike Endale

@levelsio 100%. Last time I did was December 2025

I don't write code anymore I haven't written code in I think 6 months? I think everyone is like this no?

I love good essays but some are better than others. This one is breathtakingly beautiful. Thanks for sharing, @SeanTrende

Is it only me or both Codex and Claude Code are flagging benign requests as cybersecurity risk? Got these two in 5 hours apart and couldn't get past it with any prompt change. Help? Heads up @thsottiaux @bcherny @_catwu

@thsottiaux - there might be a bug with Codex. I've been using 5.5, but the dashboard shows that I've used 5.3 spark (I'm assuming the fast mode downgrades to 5.3?) Because I ran out of 5 hour usage limit, I bought $40 worth of credits and it finished it in Standard speed in until 10 minutes. How can it use 1000 credits that quickly? Please help :)

I have no idea what @OpenAI did to Codex, but my God, it's breathtaking. It is so good right now. The elegance and reasoning process is several magnitudes better than @claudeai.

Love seeing /goal on @OpenAI Codex and @claudeai, but they both should give a nod to @GeoffreyHuntley a node for inventing Ralph Loop. He called it accurately in January that everything is a Ralph loop and it is. ghuntley.com/loop/

@ndstudio Great post! This will be really awesome if you can open source some of the evals. We’re building similar expertise in other agencies and it will be useful to learn from your patterns

How we're thinking about making AI actually useful ndstudio.gov/posts/making-a…

It’s easier to go heaven than trying to buy an iPhone from @Apple Store app. Yuck. Such a broken experience. @AppleSupport help! Please fix the app

@thsottiaux Codex needs an AskUserQuestion type of tool. Super useful to get structured questions

What are we obviously not getting right with Codex?

.@claudeai, @AnthropicAI I think users should be able to download designs that are done even after running out weekly limit. Right now, once you max out your weekly limit, it throws an "unconditional drop overload" error and chokes. Help. cc/@bcherny @_catwu

@sama @official_taches Charge $1K and give 100x the token. Heck, charge $2K if you want. Just no constraint

@official_taches what do you need for just one plan?

@alan_karthi @GoogleDeepMind @alan_karthi - this is amazing and so impactful. Our team have been working on a clinician AI tool for large Federal customer. Is there a good way to reach out? DM is not open

Since joining @GoogleDeepMind I’ve dreamt for a decade that AI can give clinicians superpowers: amplifying reach as an always-available, trustworthy member of the care team. Delighted to share our strategic research initiative at @GoogleDeepMind towards this vision: AI co-clinician deepmind.google/blog/ai-co-cli… (1/n)

This is really powerful. @GoogleDeepMind we would love to partner with you to bring this to a Federal client in the US. Is there a good way to get in touch?

@BioBayes @DrMakaryFDA @US_FDA I didn't - will do now. Thanks for sharing

@MikeEndale @DrMakaryFDA @US_FDA Very interesting!! I'm wondering, have you read these related articles? open.substack.com/pub/bessstillm…

A milestone day for clinical trial innovation. We’re announcing the first real-time clinical trials, where @US_FDA can see data signals and endpoints in real time. A quick explainer:

End of an era. Officially news from @Apple: apple.com/newsroom/2026/…

@gdb - love Codex, but the usage page is always showing 100%. I just opened the app this morning and this is what I see, but it works. Must be a bug

@patrickc Patrick - I couldn’t agree more. Helping a family member with their challenges and taking agency of their health has been a game changer. Where did you get your genome sequenced? Any reputable companies you recommend?

I'm lucky enough to have a great doctor and access to excellent Bay Area medical care. I've taken lots of standard screening tests over the years and have tried lots of "health tech" devices and tools. With all this said, by far the most useful preventative medical advice that I've ever received has come from unleashing coding agents on my genome, having them investigate my specific mutations, and having them recommend specific follow-on tests and treatments. Population averages are population averages, but we ourselves are not averages. For example, it turns out that I probably have a 30x(!) higher-than-average predisposition to melanoma. Fortunately, there are both specific supplements that help counteract the particular mutations I have, and of course I can significantly dial up my screening frequency. So, this is very useful to know. I don't know exactly how much the analysis cost, but probably less than $100. Sequencing my genome cost a few hundred dollars. (One often sees papers and articles claiming that models aren't very good at medical reasoning. These analyses are usually based on employing several-year-old models, which is a kind of ludicrous malpractice. It is true that you still have to carefully monitor the agents' reasoning, and they do on occasion jump to conclusions or skip steps, requiring some nudging and re-steering. But, overall, they are almost literally infinitely better for this kind of work than what one can otherwise obtain today.) There are still lots of questions about how this will diffuse and get adopted, but it seems very clear that medical practice is about to improve enormously. Exciting times!

@adocomplete Looks really neat, @adocomplete, but I keep getting <empty> and the process stop. Any tips?

Claude Design! It’s really good.

This greatly improves Claude Code to be more stable. I would just add auto mode by default { "permissions": { "defaultMode": "auto" }, "effortLevel": "high", "env": { "CLAUDE_CODE_DISABLE_1M_CONTEXT": "1", "CLAUDE_CODE_DISABLE_ADAPTIVE_THINKING": "1", "CLAUDE_CODE_DISABLE_AUTO_MEMORY": "1", "CLAUDE_CODE_SUBAGENT_MODEL": "sonnet" } }