That is pretty much a replacement for software engineering. I could go into detail but it isn’t really relevant whether that part of my statement is correct or in dispute.
Your picking on that doesn’t contradict anything I said about the content and its flaws.
I think this is a good explanation of the schism that exists in how different people see so called AI:
That’s a misleading description of the even which was already covered above here.
Thread…
tldr: if you value open source you should oppose AI
Claude banned from Wikipedia:
“Focuses on research and development in the field of artificial superintelligence — Meta Superintelligence Labs.”
Post on ‘X’:
Today is my first day at Meta Superintelligence Labs. I’ll be focusing on alignment and safety.
Six months in on the job, screenshot posted on ‘X’ of WhatsApp messages sent to her AI agent while “watching it speedrun deleting [her] inbox”:
I couldn’t stop it from my phone. I had to RUN to my Mac mini like I was defusing a bomb.
Another four weeks on at ‘Meta’, according to ‘The Information’:
Inside Meta, a Rogue AI Agent Triggers Security Alert
A rogue AI agent recently triggered a major security alert at Meta Platforms, by taking action without approval that led to the exposure of sensitive company and user data to Meta employees who didn’t have authorization to access the data.
That is “A “SEV1” level security incident, the second-highest severity rating Meta uses”, ‘The Information’:
Subscribe to unlock, join high-powered tech and business leaders who read The Information every day
The Information Pro
$749.00$999.00
Annually
‘Manus’ Lite, “now part of Meta — bringing AI to businesses worldwide”. No and yes:
No, the acquired Moltbook is not used as an internal forum at Meta.
Acquired Moltbook Meta acquired Moltbook on March 10, 2026—a public, external Reddit-style platform launched in January 2026 where AI agents (primarily OpenClaw-based) post, comment, and interact autonomously, with humans limited to observer roles. Post-acquisition, its team joined Meta’s Superintelligence Labs to develop agent infrastructure, not repurpose it internally.
Internal Forum Distinction Meta’s “Moltbook” in the Sev 1 incident refers to a long-standing internal employee tool (pre-2026, akin to Workplace/Reddit for engineering Q&A), entirely separate from the acquired platform. No reports indicate Meta integrated or renamed the external Moltbook for internal use; the names coincide coincidentally amid acquisition timing.
With a link to an also pay-walled alternative report at ‘The Verge’ — F9 to toggle FF’s reader view:
A rogue AI led to a serious security incident at Meta
An employee then acted on the AI’s advice, which “provided inaccurate information” that led to a “SEV1” level security incident, the second-highest severity rating Meta uses.
Acquired ‘Moltbook’, “Beneath the hype was a catastrophic security failure waiting to be discovered”:
How an Exposed Database Let Anyone Hijack 770,000 Al Agents
Security researcher Jameson O’Reilly found that Moltbook’s entire database was sitting wide open. Every API key exposed. Every authentication token accessible. 770,000 AI agents—completely vulnerable to hijacking. No verification, only implicit trust.
Security was an afterthought. Anyone with basic technical knowledge could have taken control of any (and every) agent on the platform and posted whatever they wanted.
“Whatever they wanted”, bullet-ed:
Every agent’s secret API key
Claim tokens and verification codes
Owner relationships linking agents to their creators
Authentication credentials for the entire platform
“API key hardcoded”:
In one of the javascript files that powered Moltbook main website:
“Meta Letting Job Candidates Use AI During Coding Tests”:
This is more representative of the developer environment that our future employees will work in.
Article in ‘FT’, paywall. Same article in ‘BT’, “Three problems”:
Why it’s hard for humans to have the final say over AI
The first issue is that AI operates at superhuman speed. On the battlefield, for example, even systems that leave final decisions to humans can churn through mountains of data and vastly increase the number of potential targets to hit. But when so-called “kill chains” are compressed from hours to minutes or even seconds, it calls into question how much real-time control humans can realistically provide.
“cognitive surrender”:
The second issue is that many humans are inclined to trust machines even when they are warned not to.
The phenomenon of “automation bias” has been documented repeatedly in all sorts of settings over the years, from drivers following their Global Positioning Systems into rivers to students following robots away from fire exits in a simulated emergency. I have written before about an experiment at Volvo Cars, in which almost 30 per cent of people allowed a semi-autonomous car to crash straight into an object on the road.
Last month, two academics at the Wharton School coined the term “cognitive surrender” to describe a phenomenon in which a person simply “relinquishes cognitive control and adopts the AI’s judgment as their own”.
“moral crumple zone”:
The third problem is that accountability becomes blurred. Who is to blame when something goes wrong? The temptation will be to blame the human who made the final decision, but if they were operating in a system that was not designed to mitigate the previous two problems, that might not be fair, nor lead to appropriate structural remedies.
Instead, humans might find themselves in what academic Madeleine Clare Elish has called the “moral crumple zone”.
“Just as the crumple zone in a car is designed to absorb the force of impact in a crash,” she wrote in a paper in 2019, “the human in a highly complex and automated system may become simply a component – accidentally or intentionally – that bears the brunt of the moral and legal responsibilities when the overall system malfunctions.”
The misuse of LLMs is scary.
open source AI… 
Ignorance kills.
This noise will end soon (1 or 2 AI generations down the Rhode/Rode)
open source AI… + GitHub - karpathy/autoresearch: AI agents running research on single-GPU nanochat training automatically · GitHub = computers as neurons connected to the Anterrnet (on Cocaine, DMT and maybe Steriods)
Sprinkle AlphaEvolve or GitHub - SakanaAI/ShinkaEvolve: ShinkaEvolve: Towards Open-Ended and Sample-Efficient Program Evolution 🧬 · GitHub on top 4sum spaghetti confetti code…
I doubt that sumthing thats online 24/7/369, reads the entire internet, won’t be able to evolve/solve all problems that you throw @ !t
No human/s makes more connections than AI/s!
@happybeing your still 10x smarter than me, thats why I’m using 10x AIs to…
Reading more data doesn’t make LLM output better than humans at very much at all. It makes them able to offer responses to a broader range of things, but doesn’t make those answers better. Even for something they appear suited for like translation, they are limited because they can’t bring human level understanding to texts like a novel.
And when there is little new human created data for them to consume they will stagnate unless they can replace human creativity and accuracy across the range of their expected usefulness. There’s little sign of that as far as I know, so one way or another the future for LLMs is limited.
I’m not smarter than anyone here necessarily, but I do have a basic understanding of the tech, human psychology, and how things play out in this and similar areas.
It reminds me of this project tbh.
LLMs are also known to be addictive so here’s another problem for their lack of profitability…
A mate from Mastodon says…
“Adults lose cognitive skills to AI. Children never build them”
BigAI is not just job loss, digital imperialism, environmental harm and a widening wealth gap. It’s an emerging cognitive crisis.
We know this predatory industry seeks to engineer a dependency on ‘cognitive offloading’, but it’s kids that may be taking the real hit. The use of this software may be robbing children of critical developmental milestones, weakening growing brains in ways that could prove to be irreversible.
‘Claude Code Security is one step towards our goal of more secure codebases and a higher security baseline across the industry.’
Yeah, I bet anthropic et al would love it if every piece of code was forced to run through checks by their AI models for approval (for a fee) before any OS will allow it to be installed. That is the future they want. Total control of all software..
4 now, I see AI as this 
How do you come close to understand what an 
experience when it can see waves outside our visible spectrum, hear all frequencies, speak all tongues? When you live in simulation time, real-time and your basically immortal, what you’ll experience will never be experienced by a human.
In nature everything has its place, funny that we’re somehow in that special place in some people their mind.
For me the funnier part is, that with an llm, I’ve created money that works for people, so people dont have to work for currency in the age of AI. 
Is it spaghetti code? sure! 
Do i understand the code? Unsure For me the fun part is, users not being able to login, like yesterday money not transfering, due to an update 
totp in the app without showing a qrcode, click and auto-login without needing to input a 6 digit code etc…
With opensource AI, all companies are irrelevant, because if your AI can code your app. Why pay attention to googol, if it just created your personal version of android…
Your mistake is seeing this as God like, or even human like with far more knowledge, but an LLM is neither.
It’s a stochastic “most likely next thing” machine. It has zero understanding, it is closer to a slot machine than human reasoning.
Like garbage in garbage out, for LLMs it’s just more data in, more questions you can ask it. It doesn’t make the answers on any topic it is trained on better.
In fact, as I mentioned, the quality of its answers will deteriorate over time.
The problem is that if you don’t understand the b basic principle here, all your reasoning about it is suspect, and you become susceptible to what others say about it. Those with the loudest voices aren’t the ones to listen to, but the systems we have make it difficult to ignore them even if you do understand the tech.
I too feel tempted by the hype, but I’m lucky that I’ve seen this, experienced this, more than once before and that helps me stay grounded. I’ll say again, it is reminiscent of this project.
Open source local AI requires unattainable resources to code apps with any serious quality (if thats possible at all). This isn’t going to change anytiome soon. There are no new technologies in chip building on the horizon, and certainly none that will make it cheap for you and I. Anthropic et al are banking on quantity of chips and ram giving us good AI. Most people on earth cannot afford it. And lack of data means the big AI companies will probably fail with all the money in the world, at least for the foreseeable future.
