Marsh with Water Lillies by van Gogh and from WikiArt
Really love his sketches25 March
Really interesting blog/paper about drawing:
In the paper, we argue that drawing is hard because we humans have very limited vision and memory—much more limited than we think. We just don’t perceive the world as a picture. And so, learning to draw realistic pictures is not about innate talent, it’s about learning skills to bypass these visual limitations that we all share.
So much of this tracks. I've also often thought of learning to draw as about overcoming the urge to fill in details with what you think is there, rather than what is actually there. We spend a lot of time looking people in the eye, so we think eyes are bigger and higher than they really are, etc.
25 March
Brilliant streamgraph from the New York Times on oil passing through the Strait of Hormuz
25 March
...One of the most consequential lies in Australian public life is that we are a deeply polarised nation...
But when you sit with people, night after night, and give them the space to talk about what actually matters to them, the picture that emerges is striking in its consistency. Australians want the same things. They want to be able to afford a decent life. They want healthcare and education that works. They want their kids to have a future. They want kindness and fairness. They want to know that their neighbours are OK.
Interesting post from a researcher at Redbridge
23 March
20 March
Ferrari Uncle
Just testing a new feature to include images in posts. Scribbled this at the Formula 1.
16 March
Over 10 years ago, Jazzband started as a cooperative experiment to reduce the stress of maintaining Open Source software projects. The idea was simple – everyone who joins gets access to push code, triage issues, merge pull requests. “We are all part of this.”
GitHub’s slopocalypse – the flood of AI-generated spam PRs and issues – has made Jazzband’s model of open membership and shared push access untenable.
Jazzband was designed for a world where the worst case was someone accidentally merging the wrong PR. In a world where only 1 in 10 AI-generated PRs meets project standards, where curl had to shut down its bug bounty because confirmation rates dropped below 5%, and where GitHub’s own response was a kill switch to disable pull requests entirely – an organization that gives push access to everyone who joins simply can’t operate safely anymore.
Sunsetting Jazzband
16 March
Context still matters
Nakalembe uses machine learning, computer vision, and deep learning models to map cropland, classify crop types, and estimate yields in Uganda, Kenya, Senegal, and other African nations. But most AI models are trained on European and U.S. data, and are largely useless unless they are adapted for local contexts, she told Rest of World.
“AI systems built in the West often also fail to account for the contexts of the Global South, including high internet costs, limited bandwidth, and a lack of labeled training data,” said Nakalembe, an assistant professor at the University of Maryland, and Africa program director at NASA Harvest, which uses satellite imagery to improve agricultural production.
“If these systems aren’t adapted, they remain irrelevant, potentially deepening existing inequalities in wealth and access to resources, [and] there is a risk that these systems prioritize corporate and company profit over farmers,” she said.
13 March