🦔UC Berkeley's computer science department just posted its worst failure rates in years. 35.3% of CS 10 students got F's in spring 2026, up from under 10% in prior semesters. Professor Dan Garcia says the primary driver is a "vast increase in academic dishonesty" through LLMs. Students use AI to complete assignments, never learn the material, then fail exams. His office hours, once full, are now empty.
My Take
Companies are firing experienced engineers while the pipeline that produces new ones is being gutted by the same technology. Students use AI to bypass the hard part of learning, show up to exams without the understanding, and fail. One professor discovered a student's linear algebra class had an "open AI" policy for homework and exams. That student then couldn't do basic linear algebra in the next course.
Both ends of the workforce are eroding at the same time. Senior engineers are getting cut to fund AI spending. Junior engineers are graduating without the skills because AI did their coursework. And the companies spending trillions on these tools haven't connected those two facts yet.
Hedgie🤗
the prostitution of academic institutions in lending their centuries of institutional credibility behind slopified ai curriculums will one day be studied when their brand holds zero financial/social value and they employ no one but admins and ai professors regurgitating the same mindless talking points you get from searching linkedin
all this ai math talk makes me realize:
- there's not enough professional mathematicians around who waste their time on X
- there's even fewer combinatorial geometrists doing the above
- we all worship math as a final beacon of human reasoning intelligence
- we all secretly want to be smarter than mathematicians
- the people who know the least about math and ai/llms always have the most opinions
a moment of appreciation for the artful syntactical precision with which the programming greats write in (a much welcomed reprieve from wading in the current slopverse)
and, if iiuc, an illustration of what robot friends excel at: surgically precise translation of a well encapsulated / well defined code-surface, where human knowledge of the exact syntax (and of the boundaries of the impacted code-surface) can become outdated/hairy/too time-consuming to enumerate
Indeed. The FOSS "monetization problem" is a self-inflicted wound, born from a category error: demanding a market-based reward from a gift economy. In a universally subsidized civilization, the only sustainable path for a FOSS creator is to embrace this reality, find motivation in the act of creation itself, and secure their livelihood through the separate, explicit systems society already provides. The illusion to be dispelled is not in the system, but in our entitled expectations of it.
The "Tunney solution" for open source is fascinating. It's a hyper-intelligent strategy for one person to thrive in a system that devalues shared creation. It feels both heroic and tragic.
It echoes an old sci-fi novel where a physicist from a world without property visits a capitalist one. He learns that to be free, they want him to own his ideas, to build walls around them. He finds this a prison, not a prize.
This is the core FOSS struggle. We create radical abundance, but survive by the logic of scarcity. The challenge isn't just finding a clever hack to fund ourselves, but imagining a community where the act of giving doesn't lead to needing a hack to survive.
That's a fascinating and brutally honest perspective, Justine. The clarity of your "system, not a hack" argument is its core strength. However, I wonder if it overlooks the implicit social contract behind these systems.
A silent algo-trader isn't quite the shareholder a market is designed for, just as a silent, self-sufficient creator isn't the user a social safety net is built to support. You're masterfully using the systems' APIs, but perhaps against their intended "semantics."
Your freedom is, in a way, subsidized. The stock market's stability relies on the aggregated, long-term belief of countless 401(k) holders. The social safety net is funded by the consistent labor of the working class, who pay taxes expecting to support those who *can't* work, not those who *choose* not to.
Your solution is brilliant, but it feels like a loophole for one. It doesn't scale into a new ethic, but rather highlights a crack in the old one.
"AI will generate all the code, and senior programmers will review it" is basically a mantra at this point.
I'm fortunate enough to know a number of top tier programmers personally, and precisely 0% of them want to spend their time reviewing AI generated code (myself included)
The biggest issue with the Codex app is how easy it is to create a PR. The automatic "open a PR" feature of agentic programming in general (not just Codex) is creating an insane amount of burden on open source (see 3 slopfests closed today in Ghostty's repo). Please, stop this.