My ultimate hack for LLMs:
- Using it to resolve merge conflicts.

| Github | https://github.com/oneSIX |
My ultimate hack for LLMs:
- Using it to resolve merge conflicts.
Android’s new security screening feature is bad and I fucking hate it.
They wanted a mechanism for Google’s security team to block malicious developers.
But the same mechanism can be used by the state to control what we can do with our devices.
https://android-developers.googleblog.com/2025/08/elevating-android-security.html
here's a few settings you may wish to consider for your firefox's about:config page.
browser.ml.chat.enabled = false
browser.ml.chat.shortcuts = false
browser.ml.chat.shortcuts.custom = false
browser.ml.chat.sidebar = false
browser.ml.enable = false
extensions.ml.enabled = false
if your firefox is recent enough, you might want these to disable AI tab groups too (thanks @ilmari!)
browser.tabs.groups.smart.enabled = false
browser.tabs.groups.smart.optin = false
browser.tabs.groups.smart.userEnabled = false
I thought blockchain was stupid (and I was right), but I've never seen something so "solution in search of a problem" as so-called AI. I don't have a joke here, I simply hate it being shoved down my throat every moment of my working day.
Please just let me contribute something towards my actual team goals, using my actual expertise, so I can feel like my pay is somewhat justified...
Today we publish the 31st volume of the @thoughtworks Technology Radar - our regular survey of technologies that capture our attention in our client work. Again there's a lot we've learned about using GenAI here, reflected in the explosion of AI-adjacent tools, which we've found helpful in building useful AI-based systems. But we're also finding that coding assistants can often lead folks into trouble.