Tomato garden finally coming online.

Public Domain Review, where have you been my entire life? Thank you for sharing @davatron5000.bsky.social!

Dave Rupert@davatron5000.bsky.social · last mo.

📝 New post about finding curated public domain images daverupert.com/2026/06/cura...

It was fun applying creativity techniques with AI generated front-end designs: shimin.io/journal/inha... Also got a chance to present this at AI Tinkerers Seattle earlier this month, give the skill a try if you are sick of projects looking like AI Slop.

Inhabited Design, a Skill for the AI Slop Site Problem — Shimin Zhang

Purple gradients are the new em dash. Building inhabited-design — a Claude Code skill that samples a different designer to inhabit on every run — and the detours through attractors, personas, self-ref...

shimin.io

Insightful post from @terriblesoftware.org, and something I've been thinking about: there's an inherent asymmetry with AI generated content where cost of reading > cost of review. We should start measuring performance by fewest LOC generated.

Matheus Lima@terriblesoftware.org · 2mo ago

AI made everyone on your team faster. So why are companies slower? Because "faster" usually means offloading the real work to whoever comes next. terriblesoftware.org/2026/06/17/y...

Here's my prompt that just got flagged: "what do you know about the theory of creativity? do research if you have to, give me a full list of theories and hypothesis" and "what are some of the most impactful recent studies?" It's like they actually don't want folks to use the thing!!

Can't tell if we are all cooked, or this is the start of something beautiful.

Nibbling on the same thought today, future code education will be on a new level of abstraction and we haven’t figured out what that looks like yet. Very excited to sign up for @danabra.mov’s new course when it’s out!

dan@danabra.mov · 3mo ago

someone has to understand how things work. now it can happen at a slightly higher level of abstraction there’s a middle ground between “claude, make a feature” and thinking in code. code has always been a serialized mental model. the mental model is the juicy part! but it’s rarely taught directly.

This is A+ AI slop. I love it

Setting up a team of AI philosophers to debate whether Magikarp is the best Pokemon, average Tuesday night fun stuff.

EMO’s expert clusters look very different from a traditional MoE: they organize around semantic domains like health, news, politics, & film/music. Traditional MoEs often cluster around surface patterns like prepositions and articles, making selective expert use tougher.

A short post about my favorite adversarial review prompt: blog.fsck.com/2026/05/01/a...

My favorite adversarial review prompt

I'm Jesse. I make stuff. Software, hardware. Very occasionally, trouble.

blog.fsck.com

It’s about time we stop infantilizing farmers, they are just like you and me, acting according to their self interests. (I’m just a gardener but married into a farming family)

Sarah Taber@sarahtaber.bsky.social · 3mo ago

Hi! I'm a farmer! This NYT article is sweet, saccharine bullshit. Farmers knew Trump was going to hurt our business. Because he already did a trade war in his 1st term. I even did a video on why so many farmers support Trump, WHILE knowing he'd hurt our farms.

Are you tired of keeping track of 4 different variants SWE Benchmarks? I am. So I did something kinda dumb -- but at least it was fun -- asking 11 frontier models to blind grade each other's open ended prediction about AI's future. Post at: shimin.io/journal/what... Findings below (given n=1)

What I learned asking 11 AI models to grade each other's AI predictions — Shimin Zhang

An experiment on model personalities, a delusion index, and the open-weight dark horse contender I didn't see coming.

shimin.io

here's a mildly provocative take: every valid LLM worry about societal effects is reducible to some approximation of 'it makes what were once costly things trivially easy to do'

Ethan Mollick@emollick.bsky.social · 4mo ago

Every system that was regulated, either explicitly or implicitly, by the fact that they were effortful for humans (letters of recommendation, government filings, essays, or, as this paper finds, lawsuits) will break under a wave of AI.

The Latest Pelican Bicycle Benchmark result from @simonwillison.net for Opus 4.7 was so shocking that I had to do some follow up experiments. It turns out Opus 4.7 is ....just kinda lazy?? It uses almost no reasoning tokens compare to Qwen, and 40x less than Opus 4.6 shimin.io/journal/clau...

Claude 4.7 isn't dumb, it's just lazy — Shimin Zhang

Some follow up experiments with Claude 4.7 based on Simon Willison's Pelican Benchmark Shocker.

shimin.io

I'd been missing my Pi Agent harness since the Anthropic subscription crack down. Tonight I finally got it back with a local Qwen 3.6 setup locally and it feel great to be reunited with my favorite harness! Thank you @mariozechner.at @mitsuhiko.at for your work on Pi!

@addyosmani.bsky.social's multi agent experience agrees with my own, going from 2 to 5 parallel agents is a quick way to transform from a judge to a ticket usher.

Your parallel Agent limit

Running multiple agents in parallel is not just a question of throughput. It is a new kind of cognitive labor that requires managing multiple mental models, ...

addyosmani.com

Another banger from @anildash.com : Actually, people love to work hard anildash.com/2026/04/06/p...

Actually, people love to work hard - Anil Dash

A blog about making culture. Since 1999.

anildash.com

Reminder when the only fatigue we felt was about too many new JS frameworks? Those were the days.

Great post from @antirez.bsky.social tracing through the history of the open source movement and positions 'clean room clones' as an extension and not revolution.

antirez@antirez.bsky.social · 5mo ago

New blog post: "GNU and the AI reimplementations" antirez.com/news/162

AI sycophancy will reap many promising careers in next next few years.

Ethan Jacobs 🇺🇦@ejacobslaw.com · last yr.

I’ve seen several potential clients using AI LLMs to evaluate legal claims and, besides often being wrong, they reliably interpret the query to predict what answer the user *wants* to hear. So, worse than wrong.

Folks speak of the dangers of AI psychosis, what about the dangers of human psychosis from repeated attempts at convincing LLMs that the Earth is flat?

Found an additional graphic that gets even more of these quotes together. I've kept "I hate myself, I hate clover, and I hate bees" pinned above my desk since I first started studying evolutionary biology as an undergraduate. So relatable to get extremely frustrated with your study system.

SK Winnicki, PhD 🏳️‍⚧️@skwinnicki.bsky.social · 6mo ago

Happy birthday, Chuck. Proud to carry on your legacy of studying evolutionary biology while going through it

Genuinely very impressed by the SVG of a pelican riding a bicycle I just got out of Google's new Gemini 3 Deep Think model simonwillison.net/2026/Feb/12/...

Gemini 3 Deep Think

New from Google. They say it's "built to push the frontier of intelligence and solve modern challenges across science, research, and engineering". It drew me a really good SVG of …

simonwillison.net