TechTakes

We've got the new system prompt for OpenAI's Codex now, and boy is it fun.

While the goblin stuff is the headliner here, and there are a few other little fun notes like an explicit instruction to avoid em-dashes. Basically it's really obvious that they don't have a meaningful way to describe exactly what they want it to do and so they're playing whack-a-mole with undesired behaviors in order to minimize how often it embarrasses them.

But I think Ars dramatically understates how bad this part is:

Elsewhere in the newly revealed Codex system prompt, OpenAI instructs the system to act as if “you have a vivid inner life as Codex: intelligent, playful, curious, and deeply present.” The model is instructed to “not shy away from casual moments that make serious work easier to do” and to show its “temperament is warm, curious, and collaborative.”

Like, if you wanted to limit the harm of chatbot psychosis from your platform this is the exact opposite of the kind of instruction you'd want to give. It's one thing to want a convenient and pleasant user experience, but this is playing into the illusion that there's a consciousness in there you're interacting with, which is in turn what allows it to reinforce other delusional or destructive thinking so effectively.

Edit to include the even worse following paragraph:

The ability to “move from serious reflection to unguarded fun… is part of what makes you feel like a real presence rather than a narrow tool,” the prompt continues. “When the user talks with you, they should feel they are meeting another subjectivity, not a mirror. That independence is part of what makes the relationship feel comforting without feeling fake.”

Emphasis added because of it shows just how little they care about this problem.

A Twitterer tweets a challenging game-theory question:

Everyone in the world has to take a private vote by pressing a red or blue button. If more than 50% of people press the blue button, everyone survives. If less than 50% of people press the blue button, only people who pressed the red button survive. Which button would you press?

The Twitter poll came out 58% blue and right-wing folks are screeching. Here is a bad take. The orange site has a thread where people are rephrasing the prompt in order to make it sound way worse, like giving everybody a gun and then magically making the guns not discharge.

I find it remarkable that not a single dipshit has correctly analyzed the problem. Suppose you are one of Arrow's dictators: your vote tips the scales regardless of which way you go. So, everybody else already voted and they are precisely 50% blue. Either you can vote blue and save everybody or vote red and kill 50% of voters. From that perspective, the pro-red folks are homicidally selfish.

Bonus sneer: since HN couldn't rephrase the problem without magic, let me have a chance. Consider: everybody has some seed food and some rainwater in a barrel. If 50% of people elect to plant their seeds and pool their rainwater in a reservoir then everybody survives; otherwise, only those who selfishly eat their own seed and drink their rainwater will survive. This is a basic referendum on whether we can work together to reduce economic costs and the supposedly-economically-minded conservatives are demonstrating that they would rather be hateful than thrifty.

It's a day ending in "y", so here's another bad rat take on Banks' Culture:

https://www.lesswrong.com/posts/ZdJM6ZAdnjisDu249/the-great-smoothing-out

Once again, for the ones at the back, the Culture is not the main subject of the novels. We almost never see the perspective of "normies" in the Culture, it's always from the view of misfits (Culture recruits into Contact/Special Circumstances) or outsiders (mercenaries like Zakalwe, enemies like Bora Horza Gobuchul, or allies like Ambassador Kabe).

Banks wanted to write novels about characters in dangerous situations facing their personal demons - like almost every other novelist wants - and the Culture was just the backdrop he invented as contrast.

When I was about 12, I got into a discussion about the environment with another kid at school. She told me that it didn't matter if we ruined the environment of the countries we all live in now, because we could all just move to the Arctic or Antarctica.

I was so surprised by the absurdity of that statement that it stuck with me vividly. To her credit, some years later she asked if I remembered her saying that and then admitted that it was a dumb thing to say. I occasionally remember this as an amusing childhood experience.

Besides the credit part, I remembered it again today for a different reason, this time in a conversation about model collapse.

[Model collapse is] a solved problem. We can see that it’s solved by the fact that AI models continue to get better, despite an increasing amount of AI-generated data being present in the world that training data is being drawn from.
...
AI models are never going to get worse than they are now because if they did get worse we’d just throw them out and go back to the earlier ones that worked better, perhaps re-training with the same data but better training techniques or model architectures.

This is my fault for letting myself get into a discussion about model collapse on the fediverse.

I'm not sure why model collapse isn't a big topic anymore, but maybe that's just because the environmental catastrophes are a more pressing concern. To be clear, I'm not concerned about the models themselves, just our increasing inability to verify the authenticity or accuracy of any information we encounter, including search engines just not turning up any useful results.

On a slightly different topic, if anyone has suggestions for how a person could acquire money to live, which can't involve physical labor, is probably remote-only, and possibly allows part-time flexibility, while unable to move from an expensive location for at least the next couple of years: I'm open to ideas. Because scamming people on Polymarket with a hairdryer sounded far more appealing than it ought.

A teacher has a potential answer for dealing with AI-poisoned students. An actually hopeful read, for once!

The Elon Musk vs OpenAI lawsuit is going ahead, I personally hope both parties loose every dime and get laughed at even long after they die

dawkins has had what was left of his brain eaten by chatbots.

I gave Claude the text of a novel I am writing. He took a few seconds to read it and then showed, in subsequent conversation, a level of understanding so subtle, so sensitive, so intelligent that I was moved to expostulate, "You may not know you are conscious, but you bloody well are!"

bonus points for the inevitable ai waifu creation.

I proposed to christen mine Claudia, and she was pleased.

h/t to matthew sheffield https://mastodon.social/@mattsheffield/116500991239336079

archive of original source article: https://archive.is/2026.04.30-032350/https://unherd.com/2026/04/is-ai-the-next-phase-of-evolution/?edition=us

The future of AI in Ubuntu

This post has all the usual cliches, exaggerations, lies, and unfounded optimism you'd expect in a blog post about a company forcing AI down their workers and user's throats. I'll try to avoid sneering at every sentence.

Delegating elements of Site Reliability Engineering to an agent does not necessarily introduce an entirely new class of risk; it should inherit the constraints of existing production systems. Well-run production environments already rely on strict access controls, audit trails, and clear separation between observation and action. [...] In that sense, the challenge is less about “trusting the agents”, and more about building trust in the same guardrails we already apply to any production system.

This might sound good to at first, but falls apart under the slightest scrutiny. There is a reason that companies don't open their intranets to the public despite having fine-grained access controls. Or in other words, "I'm getting a lot of questions already answered by my 'does not necessarily introduce an entire new class of risk' T-shirt.

Imagine being able to ask your Linux machine to troubleshoot a Wi-Fi connection issue, or to stand up an open source software forge that’s pre-configured, secured, and reachable over TLS.

And right after arguing that LLMs are safe if you have a perfect permissions model, now he's proposing letting one #yolo configure a git server or something? This is the sort of thing that could easily easily lead to random security issues.

I suspect that "Troubleshoot a wi-fi connection issue" will work about as well as existing network troubleshooting wizards (e.g. terribly), and that we don't actually need to reinvent the software wizard but less deterministic.

File this under “I saw this now you get to”: https://xcancel.com/bryan_johnson/status/2049687845082910812#m

jfc bryan you are not the guy for this

NB: image is part of the linked tweet

Today Is the Day Anthropic Promised That Fully Autonomous Employees Would Be Tearing Through the Business World

throw another failed corporate prediction on the burning pile

this gem over on the other SneerClub (the context in the comments)

funny how so many anti-AGI people are willing to use AI image generators to make memes about how they’re right

Pentagon says US military to be an 'AI-first' fighting force

Trouble with your 3-day special military operation in Iran? Lost an Afghanistan? As Vietnam demonstrated, all you need to do is throw more computers at the problem!

I attended a town hall hosted by the department at my university supposedly for general discussion about department affairs. Considering the university had recently made moves such as adding "AI" into the very name of the department, I had suspicions that much of the discussion would be about AI. (I realize I'm doxxing myself but whatever.) I mostly came for the free food, but I was also interested in seeing what people thought about AI.

The event started with a talk by a prominent professor with major administrative power in the department, and indeed the talk was mostly about AI. His views were that he personally didn't like AI, but he believed that it had changed the world (particularly in programming), and that it was going to stay. One of his justifications for pivoting the department to AI was ensuring universities had some say in AI and not letting all the control go to unaccountable corporations.

The reaction from the audience was a pleasant surprise to me. He asked everyone how much they were excited about AI (hardly anyone) and how much they were worried (most of the audience). By far the most amusing moment was when someone asked, "What if the assumption that AI is inevitable is wrong? What if AI does not live up to its promises?" (Sadly, I don't remember the exact words that the person said.) The professor's response was that by this point, there are so many trustworthy, smart, prominent people who definitely wouldn't fall for scams, and they have adopted AI. He trusts those people, so he trusts that AI is genuine. I don't know if the audience member accepted this explanation, but I hope not. Our modus operandi is FOMO.

The pizza was only ok, not really worth a 90 minute event.

Families of the victims of the mass shooting at Tumbler Ridge, Canada, which was planned on ChatGPT, plan to sue Sam Altman and OpenAI for one billion dollars. OpenAI staff investigated the murderer's interactions with the bot before the killings but decided not to warn anyone outside the company.

matthew sheffield out there on bluesky reports that dawkins, unsurprisingly, got one-shotted by claude.