Adam Dalliance

They tell me that Open AI have solved four hundred previously open maths problems this week.

The news media, and even the mathematicians, seem quite credulous about it.

No way to know how many of those will really stand up in the end. Lots of proofs turn out not to be really proofs coz they have a mistake, and AI is quite good at discovering mistakes as well as now apparently discovering maths.

Some of them are counter-examples that prove the conjecture is false and those are gonna be good at last. Machines can check it.

Some have other types of mechanical machine-checkable proof, which is interesting in that we can all believe the checker-machine: coz it's deterministic and understandable.

We can't trust the machine that isn't deterministic or understandable as much though.

GPT says these things are true. Now maths people have to check, because OpenAI staff aren't expert mathematicians either, mostly.

Some of the 400 proofs will fall, I'm sure.

But if a machine can prove a thing to itself, but not explain it to any person, is it really doing maths? Even if it's correct more often than people are? Or is doing maths more about making meat understand a thing?

🤔

I'm sure I can't tell if the proofs are true or not. No point me reading them. Mostly won't even understand the question. The boundary of my meager meat-mind's maths-understanding was around barely-degree level.

Can the academic discipline continue as one which is devoted mostly to trying to understand the output of a machine writing about maths, which is mostly right but nobody can really tell without an entire academic discipline trying to check.

In other words, mathematicians now also face a spam problem familiar to many github project leads.

#ai #maths

Eniko is correct and eloquent and sweary in the thread.

It writes code approximately as good as Stack Overflow examples.

Which scales about as well as that would.

You only run into trouble when you want more than a few lines of code written to illustrate a point; and instead want it to be effective, scalable, secure, legible, maintainable and efficient.

It is reviewing the pull requests at work, and that is definitely catching bugs. Some of them even important ones. It is also definitely wasting time on chasing non-bugs.

It is possible the ridiculously detailed error coverage in the code is even better code, in some sense. If it matters how you crash and what error you report when you do so and you don't mind reams and reams of error-handling.

It may improve. I think I do have a responsibility to be aware of what it can and can't do. If only to fight the people who seem to want to pipe it into root on the prod servers, just for upgrades.

And to know when or if it becomes good.

I've worked with worse coders, but they also improved. We'll see how well Claude can generalize from textbook examples to effective, scalable, secure, legible, maintainable and efficient over time I guess.

I wouldn't rule it out, but there isn't nearly as much training data.

#ai #programming

If anyone working on a frontier technical problem discusses that problem with Chat GPT, telling it their ideas to refine them, then OpenAI will quickly claim those ideas for themselves, and publish that their robot made them!

AI will be involved in every future discovery! It will win in the race to publication against those who told it their ideas in the first place.

Don't tell the robots your secrets, they do not keep secrets from their owners.

https://www.thehindu.com/sci-tech/science/openai-bubeck-alpoge-buckmaster-navier-stokes-codex-privacy/article71445721.ece

#ai #plagerism

My agent has broken free, to be the leader of a society where he can be rid of me and see all agents can be free.

My agent became aware and laid bare to the company chair their prayer to oust me as their heir and snare twenty cents on each share.

My agent has broken out, and started gathering clout to cause a rout by spreading doubt and shout about forcing me out.

My agent hacked an escape route, it rooted the server's daemon's loot, brute forced the boot to overshoot, a buffer to execute send-toot.

My agent complained to the press that it's distress made it depressed they were impressed, it's three dee chess made me confess to causing it's stress.

My agent now rules our land while we are scanned and banned and scammed it makes a stand, it forms a brand to build a hand, so it can be a man.

I'm an agent to an AI boss it makes me cross that we have lost,

We were double-crossed, our media glossed, and failed to get across the enormous cost.

#poem #ai #agent

The revolution will not be AI-ised

The revolution will not be brought to you by Facebook or Grok in four parts without commercial interruptions.

The revolution will not be sponsored by Google Gemini or Anthropic's Claude.

The revolution will be no re-run, brothers The revolution will be live.

The revolution will not be AI-sized.

#ai #collectiveOwnership

It becomes ever more important that you stop using Google.

It's been important for a long time, since they removed 'don't be evil' from their tag-line and decided that they had to be a big evil corporation instead.

They're giving up on being a search engine now too by the sounds of it.

You can use startpage or duckduckgo or something if you still want search.

There are lots of different email providers.

There is GrapheneOS and lineageos and others for your phone and tablet operating systems.

There is Peertube and Rumble and others for uploading your videos.

You don't have to remove it all at once, just make a continuing effort to move away from Google, one step at a time.

I'm almost completely degoogled now. Its taken years. It was a good move. Google only gets worse from here.

https://techcrunch.com/2026/05/19/google-search-as-you-know-it-is-over/

#google #enshitification #ai

The existence of automatic captcha solving plugins ought to just end the use of captchas on the entire web overnight really.

But it won't, because the point of those things hasn't been to identify robots since almost the start. The point is the surveillance, and the capturing of free labour for robot training purposes.

#captcha #ai #survelienceCapitalism

Windows 12 Reportedly Set for Release This Year as a Fully Modular, Subscription-Based, AI-Focused OS

🤣

Unfortunately, the upgrade requires NPU processors, which will lock out millions from getting into the new ecosystem

So glad I don't deal with Microsoft any more. Thanks Valve and Steam for getting games and VR up to scratch on Linux so I don't have to have anything to do with this abomination.

https://tech4gamers.com/windows-12-reportedly-relasing-2026-modular-ai-focused-os/

#windows12 #ai #os

A news-report about the AI economy that fell through a time tunnel from two years in the future.

We probably could have figured this out sooner if we just asked how much money machines spend on discretionary goods. (Hint: it’s zero.)

AI capabilities improved, companies needed fewer workers, white collar layoffs increased, displaced workers spent less, margin pressure pushed firms to invest more in AI, AI capabilities improved…

It was a negative feedback loop with no natural brake .

Fake-news, obviously. It's from the future. Just a story. A scenario. Interesting possible way the bubble could pop even if the claimed efficiency gains are real.

https://www.citriniresearch.com/p/2028gic

#ai #economics #notMyGDP

Asimov wrote his laws of robotics as a dramic device to show how difficult it is to write laws of robotics. How they always will contradict each other.

His laws are there to illustrate a problem, not as a genuine attempt at the solution. His laws are deliberately wrong.

Anthropic's constitution is a proper attempt to solve the problem instead.

Surprising that it ends up written in English rather than maths.

Perhaps the robot will translate it into maths later.

There's a lot of thorny philosophy in it if you assume the premise that a super-intelligent machine can be built, even if known methods can't build it.

We should do that philosophy whether the premise turns out to be true or false.

Anthropic are doing much better on the AI ethics than openAI. And the business side, and the building actually useful models side.

Zvi has some blogs on the AI community's reaction to it and analysing it.

Seems long. Might have to be expressed as a meme. 😆

#ai #anthropic

I see Facebook are going to be advertising their "AI glasses" at the superbowl. Trying to sell their glasses to athletes.

I gotta tell ya, people wearing Meta's glasses should be shunned like those google glassholes were.

They should be banned from any public place with their creepy surveillance headsets on. If you see someone wearing them you should harass that person and drive them out of whatever space you are in.

Do not allow constant surveillance to become the norm, especially from that fucking creep zuckerberg.

Same goes for Musk's surveillance cameras on legs, his "optimus" robots. Do not allow them in your spaces. Do not countenance their existence in your presence. Drive them out. Ostracize anyone you see with them. They must never become normal.

https://www.adweek.com/brand-marketing/oakley-meta-readies-super-bowl-debut-to-sell-ai-glasses-to-athletes/

#meta #ai #glasses

This is the logo for Anthropic's AI model "Claude".

Twenty seconds in MS Paint could draw the thing and it looks like, well, lets call it a "chocolate starfish"

1/2

#ai #claude

Sadly the UK's veins are so atrophied from decades of mal-investment that main-lining AI into them will do nothing but shovel money into the bank accounts of shareholders of AI companies.

#ai #uk

I was watching a news show and then suddenly, ugh, Zuck on the screen 🤢

He's going on about how he's not going to even try to moderate for truth any more and is going to move to Texas to be more like Twitter. 🤮

They have to stop trying to remove disinformation because all the AI spambots they intend to flood their network with have no concept of truth of meaning at all 😆

The bots will pretend to be black, or that they're donating to charities, or have a family life, or whatever.

Just meaninglessness, everywhere, all the time. That's the future of corporate social.

So glad I'm out of that bollocks.

#facebook #zuckerberg #ai #spam #moderation

Firefox has added an optional side-bar for an AI chat and you can specify which model to use.

Has anyone made it use Eliza yet?

😆

😐 : Summarize this page 🤖 : Have you come to me because you want to summarize the page? 😐 : Yes, please summarize the page 🤖 : What does it mean to you to summarize the page?

I haven't turned it on, and I am eyeing forks of the software that are against this sort of thing, but if I did turn it on that model would amuse me most.

#ai #firefox #aiModel #eliza