An Autopsy of AI-Generated 3D Slop (via HN)

Handcrafted 3D relies on “edge flow,” which are lines of geometry that naturally follow the contours of the object. This allows for smooth reflections and easy editing. AI models generate meshes using “isosurface extraction” or similar volume-to-mesh techniques. This results in what the industry calls triangle soup.

The Misuses of the University (via HN)

He wonders who will train the next generation of scholars if America’s First Research University won’t anymore. Hopkins isn’t alone; many of the country’s other universities are doing the same. Maybe the leaders of our nation’s wealthiest universities can’t think a generation ahead anymore, beholden as they are to board members who think in terms of the next quarter. Or maybe they have simply given up on the future of an industry that now relies mostly on underpaid gig labor to teach students.

Practical Decentralization, Paul Frazee (via HN)

Info civics is a practice of tensions:

On the one hand, you need to achieve your ideological aims; you need to create a protocol that guarantees individual and community rights.

On the other hand, you need to produce usable software; you need it to perform well, fit the use case, and resolve problems as they arise.

The Slow Death of the Power User (Probably written by AI via HN)

The average person who grew up with smartphones has a fundamentally broken mental model of computing. Not broken in the sense that they can’t operate their devices — they can, with terrifying efficiency. Broken in the sense that their understanding stops at the glass. They know how to use apps. They do not know what apps are. They know files exist somewhere, in the cloud maybe, or possibly inside the app itself — the distinction isn’t clear to them and they’ve never needed it to be.

Puts into place an oppositional argument of why GUIs are not the best thing. Even though in the past I’ve said that graphical-user-interfaces are necessary for wide adoption of a piece of software.

Elsevier shuts down its finance journal citation cartel, Hacker News

The state of research is dire at the moment. The whole ecosystem is cooked. Reproducibility is non-existent. This obvious cartel is a symptom and there should be exemplary punishment.

Publishers are commercially incentivized to simply maximize profit and engagement. The main actors are academics and most of them try to uphold the high standards and ethics. Yes there is free-riding, backstabbing and a lot of politics but there is also reputation and honesty.

Paper idea for anyone interested “Reproducibility as metric for success in academic research” - an alternative to citations, peer-review, and other reward schemes that favor quantity over quality