>ChatGPT (o3): Scored 136 on the Mensa Norway IQ test in April 2025
If you don't want to believe it, you need to change the goal posts; Create a test for intelligence that we can pass better than AI.. since AI is also better at creating test than us maybe we could ask AI to do it, hang on..
>Is there a test that in some way measures intelligence, but that humans generally test better than AI?
Answer:Thinking, Something went wrong and an AI response wasn't generated.
Edit, i managed to get one to answer me; the Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI). Created by AI researcher François Chollet, this test consists of visual puzzles that require inferring a rule from a few examples and applying it to a new situation.
So we do have A test which is specifically designed for us to pass and AI to fail, where we can currently pass better than AI... hurrah we're smarter!
The validity of IQ tests as a measure of broad intelligence has been in question for far longer than LLMs have existed. And if it’s not a proper test for humans, it’s not a proper test to compare humans to anything else, be it LLMs or chimps.
To be intelligent is to realise that any test for intelligence is at best a proxy for some parts of it. There's no objective way to measure intelligence as a whole, we can't even objectively define intelligence.
I believe intelligence is difficult to pin down in words but easy to spot intuitively - and so are deltas in intelligence.
E.g watch a Steve jobs interview and a Sam Altman one (at the same age). The difference in the mode of articulation, simplicity in communication, obsession over details etc are huge. This is what superior intelligence to me looks like - you know it when you see it.
>Create a test for intelligence that we can pass better than AI
Easy? The best LLMs score 40% on Butter-Bench [1],
while the mean human score is 95%. LLMs struggled the most with multi-step
spatial planning and social understanding.
That is really interesting; Though i suspect its just a effect of differing training data, humans are to a larger degree trained on spacial data, while LLMs are trained to a larger degree on raw information and text.
Still it may be lasting limitation if robotics don't catch up to AI anytime soon.
Don't know what to make of the Safety Risks test, threatening to power down AI in order to manipulate it, and most act like we would and comply. fascinating.
>humans are to a larger degree trained on spacial data
you must be completely LLMheaded to say something like that, lol
humans are not trained on spacial data, they are living in the world. humans are very much diffent from silicone chips, and human learning is on another magnitude of complexity compared to a large language model training
Does an LLM scoring well on the Mensa test translate to it doing excellent and factual police reporting? It is probably not true of humans doing well on the Mensa, why would it be true of an LLM?
We should probably rigorously verify that, for a role that itself is about rigorous verification without reasonable doubt.
I can immediately, and reasonably, doubt the output of an LLM, pending verification.
The Montreal protocol (1987) put us back into the dark ages with coolants for a while (both with CFC ban and later phase outs of HFCs). I suspect if you tested a refrigerator from 40 years ago they would give modern ones a run for their money...
It was obviously a worthwhile sacrifice for the ozone layer though.
It is a criminal offense in the UK to use insulting words in public, or to send any message online that anyone could find insulting or offensive (whether any one does or not is irreverent).
The Online Safety Act and Hate Crime Provision have extended these somewhat into the realms of 1984. But the police do tend to use them sparingly.
> It is a criminal offense in the UK to use insulting words in public, or to send any message online that anyone could find insulting or offensive (whether any one does or not is irreverent).
You cannot be arrested for sending “any message online that anyone could find insulting or offensive”. That’s not what the law says. You can be arrested for spreading hate speech, inciting violence, sending illegal media or harassment online.
All of the arrests mentioned in this thread in relation to these acts have been campaigns of intimidation, harassment and calls to violence, not simply saying something “insulting or offensive”.
In the UK political expression of free speech is protected by the ECHR, which overrides both those acts (look carefully who wishes to abolish the ECHR).
> All of the arrests mentioned in this thread in relation to these acts have been campaigns of intimidation, harassment and calls to violence, not simply saying something “insulting or offensive”
This is false. But even if it weren’t, it would be unjust. Determinations like “hate speech” are subjective, and have no place in law concerning speech. Without free speech, there is no democracy.
There’s a big difference between being free to criticise the government and those who define and enforce laws, and being free to say anything to or about another citizen without repercussion, even if it may cause them harm.
The people mentioned here who were arrested due to violations of the communications acts are definitely the latter. The people arrested in peaceful protests for being associated with Palestine Action or Just Stop Oil are the former.
>In the UK political expression of free speech is protected by the ECHR, which overrides both those acts
This is categorically untrue. Not only is the ECHR worded specifically to allow individual countries to curtail free speech ("any law, deemed by the local democratically elected government as ; necessary in a democratic society, and for a legitimate aim"), but parliament always had sovereignty to pass into law exemptions to the ECHR, which we have done on multiple occasions.
Yes, this is why the government needed to label Palestine Action as a terrorist organisation. It needed special measures because it did not in fact have the authority to arrest protestors, even though some people found what they were saying offensive.
The police are overreaching massively. They are making 30 arrests a day and "interview" many more.
We do not rely on the ECHR to protect our free speech. If we did the UK would no longer be a democracy. I'm offended by the suggestion that our democracy and society is so fragile that without them we would have no rights. Expect a police raid very soon.
I think you're being disingenous. There is clearly an unprecedented and systemic effort to police social media. Even if the posts did actually violate the law doesn't change my point or address my concerns. This is not what the police should be doing.
I honestly don't know if that's true or not. But I haven't seen any compelling evidence to support it. The figures being lobbed around by the likes of Tommy Robinson are deeply harmful to the debate because they are both a) completely wrong and b) misleadingly quoted. You can lookt the actual stats here: https://www.met.police.uk/foi-ai/metropolitan-police/disclos...
We're talking on the order of a few hundred arrests per year for section 127 of the Communications Act and 1500 per year for the Malicious Communications Act, which includes stuff like racial harassment, domestic abuse, pedophilic grooming, and a whole host of things that I would hope you agree should be illegal.
The latter part at least is true. Sending "grossly offensive" messages is illegal under the Malicious Communications Act 1988 and the Communications Act 2003, specifically Section 127:
> a person is guilty of an offence if he—
> (a)sends by means of a public electronic communications network a message or other matter that is grossly offensive or of an indecent, obscene or menacing character; or
> (b)causes any such message or matter to be so sent.
I suspect the former is also true, but am not well-read in that area
If this concerns you I would advise not looking into pretty much any UK law which is full of subjective terms and ways to interpret them. The law isn’t an algorithm nor should it be. Just because you can’t understand how it works doesn’t mean it doesn’t work.
Sorry, that was a low shot, and not meant truthfully, but I just couldn’t resist making an offensive comment when the topic was “you can be arrested for offending someone”.
There's no value in making insults for the sake of being insulting protected speech, but in the UK if you're making ECHR Article 10 protected speech that someone happens to find insulting or offensive then that's not a crime. It's unhelpful to permit insults as free speech to prevent some hypothetical future abuse, since all modern dictatorships pay lip service to free speech and instead lock up their political opponents for embezzlement or mortgage fraud or whatever.
Humans don't think. At all. They do next token prediction.
If they are [raised in environments] that includes lots of examples of the result of people thinking, what they produce will look sort of like the results of people thinking, but then if they were [raised in an environment] of people repeating the same seven knock knock jokes over and over and over in some complex pattern (e.g. every third time, in French), what they produced will look like that, and nothing like thinking.
I believe this can be observed in examples of feral children and accidental social isolation in childhood. It also explains the slow start but nearly exponential growth of knowledge within the history of human civilization.
I’m not going to hash out childhood development here because I’m not paid to post but if anyone read the above and was even slightly convinced I implore you to go read up on even the basics of early childhood development.
> I implore you to go read up on even the basics of early childhood development.
That's kind of like taking driving lessons in order to fix an engine. 'Early childhood development' is an emergent property of what could be cumulatively called a data set (everything the child has been exposed to).
The neurons in a child's brain might be 'wired' to accept data sets, but that does not make them fundamentally different from AI systems.
Are you claiming that a child who is not exposed to 'reason' will reason as well and one who is? Or a child who is not exposed to 'math' will spontaneously write a proof? Or a child not exposed to English will just start speaking it?
01101100 01100101 01100001 01110010 01101110 may be baked into US and AI in different ways but it is fundamentally the same goal and our results are similarly emergent from the process.
>What is so special about services delivered over the internet?
The most dangerous people on earth who are not in prison are on the internet; It is an adult place. Making it look like a child friendly place will not change this. But it will lure more kids online unsupervised and unprotected.
I purchased one for £24, 11 years ago, and still use it. Ironically from amazon (code B006GTOYDS) - can't find it in way back machine still have it in my purchase history. Before they started killing the competition. Miles better than the kindle to read in daylight and battery life still lasts weeks... cheap isn't always throwaway.
In contrast, i know people who have went through many kindles in this time and spent a small fortune on them.
> Miles better than the kindle to read in daylight
Are you comparing it to the Kindle ereaders or their tablets? Standard (non-tablet) Kindles such as the Paperwhite series are like you describe (though they cost more than $100 and come with all of the lockin issues).
I don't think they would have a choice. In a scheme like this, publishers would lock it down so they can sell it to you again. Perhaps even only allow a particular book to be read.
Because they could always buy a replacement for cheap. Sadly, our consumer goods prices do not correctly reflect environmental externalities, so in a way a higher price is better for the environment, even if the difference doesn't go in the right pocket.
So if your training and double your water intake your basically lowering you IQ? (according to the Chinese studies) I wonder the method this uses.. has anyone looked at dementia rates in high fluoride areas.. Particularly in people with high water intake?
There is also a host of things we use water for from cooking to preserving, distilling and cooling.. i wonder if any of these things could concentrate the fluoride.
Also since fluoride has a lower boiling point any studies tracked what breathing in fluoride gas over long periods cause?
When you are released from prison, they can simple ask you to decrypt the data again, and if you refuse or can't, you have broken a law with another 2 years in prison (5 if they think you could have anything to do with 'terrorism').. Its theoretically an infinite prison sentence for forgetting your passwords.
Sounds like one of these stats where they just invert the cause and effect to get a story; i.e. People who are healing better will obviously walk sooner. Inverted to people who walk sooner are healing better.
There is a lot of hokum and bad statistics in the medical field. Doctors truly don't have a great idea what improves post op outcomes.
There are some bigger studies coming out that show that early weight bearing is non-inferior to traditional protocols that ask for many weeks of NWB though, and given the obvious qol benefits of walking earlier it seems to me the standard should be mobilise ASAP.
There really isn't good evidence for immobilisation. It seems to be a hold over particularly for surgical fixation, where there's no real fear of displacing things if it's been fixated properly.
If you don't want to believe it, you need to change the goal posts; Create a test for intelligence that we can pass better than AI.. since AI is also better at creating test than us maybe we could ask AI to do it, hang on..
>Is there a test that in some way measures intelligence, but that humans generally test better than AI?
Answer:Thinking, Something went wrong and an AI response wasn't generated.
Edit, i managed to get one to answer me; the Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI). Created by AI researcher François Chollet, this test consists of visual puzzles that require inferring a rule from a few examples and applying it to a new situation.
So we do have A test which is specifically designed for us to pass and AI to fail, where we can currently pass better than AI... hurrah we're smarter!
reply