Oleh Aleinyk
@oaleinyk.bsky.social
Software engineer, bookworm, 🇺🇦 oaleinyk.xyz
A lot of posts lately about Anthropic starting to watermark text and people worrying about it. Google has been doing that with Gemini since 2024 deepmind.google/blog/waterma...
Watermarking AI-generated text and video with SynthID
Announcing our novel watermarking method for AI-generated text and video, and how we’re bringing SynthID to key Google products
deepmind.google
Heartbreaking. And a wake-up call. It is no longer about being difficult to find a way to sustain yourself through your work online. It is no longer sustainable to keep a website up now. Axel has done so much for the web and all of us. Buy one of his books if you can. We've learned so much from him.
Axel Rauschmayer (also on 🦣)@dr-axel.de · 4mo ago1/ I’m still processing what’s going on in tech at the moment. I used to feel at home in that world. It used to feel incredibly creative and positive. Now it’s increasingly full of bugs, dark addicting patterns and AI slop. 2ality.com
We need everyone to wake up to how fragile our shared commons are. We can't wait until we drown in security incidents and more projects close down, using them as excuses. If you're in a position to help (every company is), check who you depend on and need support (everyone), and fund them.
I just saw the mention of the HTTP/S proxy agent and had to pause for a moment.
I wrote about why I quit my job, and how weird and tiring tech feels these days.
Do I belong in tech anymore?
On quitting, the spread of AI, and the loss of an ideal.
ky.fyi
jj workspaces feel much nicer for parallel work than Git worktrees, and the important part is not the shared repo. In jj I can just start working in a workspace, and the work becomes part of history through @ without first creating a branch or deciding to commit.
Do you ever clean up your old GitHub repos and forks? 🧹 Wondering if anyone actually bothers or just lets them pile up.
Oh, the luxury of just finding a calm place and reading a book. Like a lotion for an irritated mind.
Rewrote my agent coding skills from checklists to thinking instructions. Specific rules like "use X not Y" ground the agent and go stale as the language evolves. Guiding it to ask "is there a simpler way?" and self-research before deciding lets it go much further. Cognitive quality instructions?
One thing I'm realizing with code agents: the speed gain shouldn't become pressure to ship more. I want to use that time for quality. Better tests, thoughtful design, simplification. Boring is the goal.
Reflecting on a pattern: each time we adopt new tools, we discover they weren't built for the new intermediary — search engines, app stores, algorithms — and rush to optimize. The product drifts a little further from the person it was meant to serve. "AI-first" codebases feel next in line.
I've gotten more from practical experience than from memorizing "best practices" — and that goes for prompting AI too. Concrete examples show you the tradeoffs. Best practices usually skip that part. In practice, "when does this apply?" gets me further than arguing about whether it's the right way.
Raw skill and knowledge are more important than ever before.
What if MCP had the same kind of progressive disclosure as skills? 🤔
One useful addition to agents.md I made to better align CLI agents with the “reality” is: “use `date` command to get my current date and time when you need to work with dates.”
Andy exemplifies effective web practices and design. Highly recommend the courses offered by Andy and his team.
Andy Bell@bell.bz · 10mo agoA bit of a vulnerable post by me but we need your help to get us through what has been a more difficult period than 2020. It would be really appreciated if you could read and share in your networks too 💛 bell.bz/its-been-a-v...
Really digging the workflows the Wispr Flow provides. Added their blog to the list of favourites. Their approach reminded me a lot of this blog post wattenberger.com/thoughts/our... by @wattenberger.com
What are the good starting points to start hacking with atproto? 🤔
Noticed that amp became much more expensive (MCP might be contributing here, not sure 🧐). Tried their race mode for a change (haiku 4.5) - super fast, results are very close to what I need most of the time, and it is cheaper.
`amp --ide` + neovim amp plugin + custom commands—just enough AI code assistance for me 👌 Excellent, pragmatic solution from @sourcegraph.com
we now have some new @e18e.dev docs on best practice of publishing npm packages this documents the recommended basics for a secure publish workflow and gives some pointers for further security/tools/etc this is a _very_ opinionated subject, so do ping me if you have feedback!
e18e (Ecosystem Performance) - Publishing Packages
Best practices on publishing npm packages securely using GitHub Actions.
e18e.dev
Dynamic Worker Loading is super intriguing. I can see how this can be used for something like “morphing” APIs, kinda GraphQL on steroids. I wonder about the potential of this in workflows… Thanks for sharing and working on this @kentonvarda.com and @threepointone.bsky.social 🙌🏻
Kenton Varda@kentonvarda.com · 12mo agoThis is all based on a new Workers Runtime feature called Dynamic Worker Loading, which lets you dynamically load arbitrary code into a fresh isolate. Isolates are wayyyyyy cheaper and faster to load than containers. Perfect for executing one-offs. developers.cloudflare.com/workers/runt...
This is a piece of art from @wattenberger.com. I stayed there for more than 30 mins, carefully scrolling and checking the details of each image. I practice something similar with sound and LLMs, but that example takes multi-modal system interaction to the next level. wattenberger.com/thoughts/our...
Our interfaces have lost their senses
wattenberger.com
Oooh, I totally love this! In iOS 26, they've added some cool new background sounds.
