Most projects don't have enough people or external interest for that kind of process.
It would be possible to establish some tooling like that, but standard forges don't provide that. So it'd feel cumbersome.
And in the end you're back at having contributors, trustworthiness, and quality control. Because testing and reviewing are contributions too. You don't want just a popularity contest (I want this) nor blindly trust unknown contribute.
The because of training claim is wrong.
Quoting the Gentoo post:
Mostly because of the continuous attempts to force Copilot usage for our repositories,
It seems to be about GitHub pushing copilot usage, not them training on data. Moving away doesn't prevent training anyway. And I'm sure someone will host a mirror on hitting if they don't.
Or maybe the bubble will pop, and we’ll have a repeat of 2008, except 100 times worse.
The AI spending and stock is much more localized. It's not grown over more years and integrated into many financial systems. So the results may be much less severe.
I've always bought Western Digital. The one time I bought Seagate I eventually had data loss. Stuck to WD ever since.
I'm not sure feeding more misinformation to our systems and society is that good of an idea. I don't think it'd be an effective influencing strategy either.
Additionally, patentable materials must be novel, useful, and a non-obvious inventive step. - Wikipedia
What does the patent contain? Where is the non-obvious inventive step? Using an AI to impersonate someone doesn't strike me as novel, inventive, or surprising.
From The Software Quality and Productivity Crisis Executives Won’t Address (via on Lemmy)
Executives aren’t ignorant. They have the data. They commission the surveys. They attend the conferences where CTOs present their concerns. They know that:
- 91% of CTOs cite technical debt as the biggest challenge
- 75% of projects are expected to fail
- 69% of developers lose significant time to inefficiencies
- Only 39% of projects meet success criteria
- The recommended 15–20% investment in technical debt management yields better long-term returns than crisis spending
Yet they choose:
- Not to allocate recommended budgets for technical debt management
- Not to make quality a strategic priority despite CTOs’ and developers’ concerns
- Not to mention these challenges in public communications to shareholders
- To celebrate AI productivity gains whilst developers report record inefficiency
- To focus on the next hype cycle (AI) rather than address fundamental problems
This isn’t a failure of knowledge. It looks to me like a failure of courage and integrity. A failure of the very concept of leadership.
If it's only vibe coding, going by vibe, being an expert makes no difference.
If you go for thorough reviews and corrections, supposed code generation efficiency gains are typically offset by review and correction effort.
function GetRandomNumber(): return 42;
It's a random number, what more do you want 🤷♂️ /s
An LLM wouldn't have made that mistake. /s
Prompt: On line 23, between the words fn and close, write void.
What do you man by tell it to write good code?
You mean specific prompts, which you mention afterwards, but also hit and miss?