In case you missed it, Dave Farley wrote an article on #AI replacing software engineers. www.linkedin.com/pul...
A dev team saying they're "done" because they shipped a feature is like a doctor saying they're "done" because they gave the patient the medicine. We're not done until we've solved the problem (or given up trying). This, I regret to inform you, requires you know what problem you're trying to solve.
Allen Holub@allenholub.bsky.social · 3mo agoMy measure of productivity is the time it takes to get the solution to a user's problem (their story) into their hands. I couldn't care less about vanity metrics like output volume or token counts. What matters is delivering value sooner. 1/2
There is something brutally psychically injurious to being constantly told "just do good work and you'll be successful" while watching people who do the most dogshit work possible do all the succeeding.
I find the acceptance and defense of mediocrity to be deeply depressing. This is true not just in the context of large language models, but particularly in the case of certain scholars and politicians.
"Once men turned their thinking over to machines in the hope that this would set them free." "But that only permitted other men with machines to enslave them." --Frank Herbert, Dune
I don't really post things on the internet much anymore because my industry and the world at large have both gone mad. I'm just a bit shocked/stunned and don't really know if the world needs me to add my word noise into the LLM-ridden husk that remains of what was formerly the internet.
And what do we call a thing that is expressive and precise and deterministic and repeatable so as to produce reliable executable artifacts? We call it a programming language. slashdot.org/story/26/03/...
Will AI Force Source Code to Evolve - Or Make it Extinct? - Slashdot
Will there be an AI-optimized programming language at the expense of human readability? There's now been experiments with minimizing tokens for "LLM efficiency, without any concern for how it would se...
slashdot.org
Billionaires are morons who got lucky. They have no better idea what will work than anybody else. See also: every overpaid CEO.
Dave Karpf@davekarpf.bsky.social · 6mo agoMark Zuckerberg, 2021: “within five years, [Facebook] will be known primarily as a metaverse company.” Mark Zuckerberg, now:
Great reporting from @rafeuddin.ft.com on Amazon's misadventures with AI coding (which have coincided with big headcount reductions) www.ft.com/content/7cab...
We don’t talk enough about how morally depraved the tech industry turned out to be. Every single ounce of their self-regarding statements of values was an outright lie.
"Developers won't need to understand syntax anymore, just as long as they can <goes on to describe things that will require understanding of syntax>"
I see posts on LinkedIn from people who believe dev teams should be pushed to adopt "agentic" coding so they can reap the "10x" productivity gains. Putting aside the lack of evidence for even 1.5x gains on teams, if it really were possible, why would they need to be pushed towards it?
If I could pick one word to sum up the last decade, it would be "preoccupied". When can we start getting on with our lives again?
a man in a plaid shirt and tie is typing on a keyboard with flames coming out of it
ALT: a man in a plaid shirt and tie is typing on a keyboard with flames coming out of it
media.tenor.com
Ten to twelve years from now wars will be fought over the remaining sources of fresh water because of these fucking dipshits.
Nathan Kalman-Lamb@nkalamb.bsky.social · 7mo agothe AI evangelists are not okay
Carl Sagan's The Demon-Haunted World should be required reading in high school. And congress.
Technical debt is organizational debt wearing a compiler-approved disguise.
A new study from Anthropic finds that gains in coding efficiency when relying on AI assistance did did not meet statistical significance; AI use noticeably degraded programmers’ understanding of what they were doing. Incredible.
Pavel@spavel.bsky.social · 7mo agoThe latest from Anthropic: using Anthropic's products makes you worse at your job
Seeing more and more posts from engineering leaders along the lines of "Tried Claude Code over the Xmas holiday and generated 10 squillion lines of code (that almost worked) in 10 minutes. Now mandating all our engineers use it as much as possible. Because I don't understand bottlenecks."
Very often when trying to think about a problem I start wondering: "How does software X deal with that?" And 98.5% the answer is: they don't, they just let it fail. Often silently. Every time I keep relearning how low the standards usually are in this industry.
The software industry's lack of concern for business or user outcomes has never been more visible.
I really do love the spec style syntax for tests (describe, before each, it). When used correctly it's the best at representing test setups and expressing test intent. I also recognize it has many downsides that have to be accounted for. Let me tell you, it really sucks resolving merge conflicts.
And, of course, there'll be the folks who warn about mythical developers "wasting time" making the software "too good".
As our civilisation relies ever more critically on software, we've collectively decided this would be a good time to lower our standards?
You call it "over-engineering" when devs make software more complicated than it needs to be. But simpler solutions often require *more* thought. Complexity's easy. You just keep typing. That's why I call over-complicating "under-engineering".