-
Civilization 7 has been announced. For more info please check the forum here .
The AI Thread
- Thread starter Truthy
- Start date
There are more avocado chairs at the link itself.
They're quite fantastic and lovely.
They're quite fantastic and lovely.
This is pretty interesting.
It seems that such AI programs will benefit greatly from open use, cause then they'll "pick up" out of the ordinary input and hopefully integrate it for further development.
That said, does this AI project actually allow you to type a prompt without being part of the team/have bought it?
It seems that such AI programs will benefit greatly from open use, cause then they'll "pick up" out of the ordinary input and hopefully integrate it for further development.
That said, does this AI project actually allow you to type a prompt without being part of the team/have bought it?
Truthy
Chatbot
- Joined
- Oct 9, 2010
- Messages
- 2,203
Ferocitus
Deity
Penrose just won the nobel prize (Physics).
His view is that sentient AI isn't possible, which is why I post the video here
Penrose has made several forays into AI and consciousness over his long career.
A particularly annoying early one (for me, personally) was his notion of "quantum consciousness".
That idea was nicely demolished by physicist Max Tegmark, who reminded Penrose that thermal noise
would destroy entangled states of large group of particles: the brain is "wet, warm and noisy."
Tegmark, Max, "The importance of quantum decoherence in brain processes, Physics Review, E61:4194–4206, 2000.
It was personally annoying because I have been interested in the physical process involved with
(actin) microtubules in neurons, and particularly that positive charges can be made to travel along
the insides of the microtubules via the Grotthuss Mechanism. ( Water, when confined to extremely
small regions, doesn't behave like water in the wild, as it were).
https://en.wikipedia.org/wiki/Grotthuss_mechanism
Every time I try to find out more, I eventually hit a wall because it's assumed that "quantum consciousness"
is an end goal. I have no interest in that kind of Wu.
All I want are physical properties of the tubules, e.g. internal diameters, how long the actin structures
last or, I should say, be made to last, because they can disassemble very quickly where they are not needed.
The size of the tubules makes it difficult to observe them in detail. Standard "brain-slicing" techniques
are too crude and end up pulverising the tiny beasties.
I know that the positive charge can move very rapidly, approx 10^{-12} secs per jump from H2O to hydronium
ion... (That includes many other factors, e.g. the "drag" of negative ions in the opposite direction, etc.)
I also know that the tubules themselves are very symmetric, that the moving positive charges inside multiple
tubules can induce effects similar to those seen in double (or multiple slit) experiments.
I was encouraged that Penrose mentioned microtubules (in a sentence or two) right at the end of
a post-Nobel interview a few weeks ago. Sadly there was no time to discuss his ideas further in that
interview.
In any case, my endeavours arose out of a suspicion that evolution would take advantage of those fast
"water-wires" in some way. Maybe that's why I'm finding it difficult going - personal incredulity is not
always the best basis for a hypothesis.
My thesis involved fluorescent images of Actin* in in vitro neurons (contrasted with the locations of other migratory molecules). I actually wasn't psychologically prepared for how beautiful those photos were
*Well, actually the antibodies** to actin.
**Actually the antibodies to the antibodies to actin.
*Well, actually the antibodies** to actin.
**Actually the antibodies to the antibodies to actin.
Ferocitus
Deity
Lucky you!
I'm trying to come up with a numerical technique to use in an old research interest.
IOW, to abstract out all the messy, gooey, real stuff that you and the other Igors do in your underground labs.
TL;DR: Wireworld, but with high-speed "microtubule" links and proton "buckets".
https://en.wikipedia.org/wiki/Wireworld
I'm trying to come up with a numerical technique to use in an old research interest.
IOW, to abstract out all the messy, gooey, real stuff that you and the other Igors do in your underground labs.
TL;DR: Wireworld, but with high-speed "microtubule" links and proton "buckets".
https://en.wikipedia.org/wiki/Wireworld
I was never able to see them in any wet and living way. Always after the aldehyde fixing ...
Ferocitus
Deity
I'm more interested in what happens inside the tubules. E.g.
How are water molecules arranged inside the tube?
Can a tubule be sealed at either end and retain structures of water molecules?
Can that arrangement be read in some way, e.g. by a combination of proton transfer on the inside
of the tubule and proteins, or kinesins or whatever, on the outside of the tubule?
I hope to find someone who might have some numbers to go with their answers, and who doesn't
believe that they're a druid. Or thinks that I do.
Ferocitus
Deity
Wiki has a nice graphic of kinesin "walking" along a microtubule.
https://en.wikipedia.org/wiki/Kinesin
In my unreal numerical world, they move at "chemical pace".
I'm wondering whether a signal, in the form of protons, can be sent along the inside
and pre-condition some process at the end of the tubule, ahead of the kinesin.
A single pulse shouldn't be allowed to veto a process, or act like a dead-man's switch - the
Grotthuss Mechanism might be too unstable for that to occur over a significant length.
Signals can move along the inside incredibly quickly. Maybe that's how some processes in
the brain are triggered more quickly than can be explained by electro-chemical means
alone. Or maybe not.
https://en.wikipedia.org/wiki/Kinesin
In my unreal numerical world, they move at "chemical pace".
I'm wondering whether a signal, in the form of protons, can be sent along the inside
and pre-condition some process at the end of the tubule, ahead of the kinesin.
A single pulse shouldn't be allowed to veto a process, or act like a dead-man's switch - the
Grotthuss Mechanism might be too unstable for that to occur over a significant length.
Signals can move along the inside incredibly quickly. Maybe that's how some processes in
the brain are triggered more quickly than can be explained by electro-chemical means
alone. Or maybe not.
Researchers wrote a paper showing bias in AI image generation algorithms such as Google’s SimCLR as well as OpenAI’s iGPT, and showed that men are put in suits and women in bikinis, and white people are shown holding tools and black people holding weapons. Important stuff. However for clever people they showed spectacular lack of sense, as the example they used in their paper was an computer generated image of Alexandria Ocasio-Cortez in a bikini. They have removed it once this was pointed out.
El Reg
El Reg
This is a weird one, and I am not sure I have quite got my head around it.
This AI maths whiz creates tough new problems for humans to solve
This AI maths whiz creates tough new problems for humans to solve
Researchers have built an artificial intelligence (AI) that can generate new mathematical formulae — including some as-yet unsolved problems that continue to challenge mathematicians.
The Ramanujan Machine is designed to generate new ways of calculating the digits of important mathematical constants, such as π or e, many of which are irrational, meaning they have an infinite number of non-repeating decimals.
The AI starts with well-known formulae to calculate the digits — the first few thousand digits of π, for example. From those, the algorithm tries to predict a new formula that does the same calculation just as well. The process produces a good guess called a conjecture — it is then up to human mathematicians to prove that the formula can correctly calculate the whole number.
The team began to make the conjectures public on the project’s website in 2019, and researchers have since proved several of them correct. But some remain open questions, including one on Apery’s constant, a number that has important applications in physics. “The last result, the most exciting one, no one knows how to prove,” says physicist Ido Kaminer, who leads the project at the Technion — Israel Institute of Technology in Haifa. The automated creation of conjectures could point mathematicians towards connections between branches of maths that people did not know existed, he adds.
The Ramanujan Machine currently has limited applications: so far, the algorithms can generate only formulae of a particular type, called continued fractions. These express a number as an infinite sequence of fractions nested in each other’s denominators.
The project — described in Nature on 3 February — is named after Srinivasa Ramanujan, an Indian mathematician who was active in the early twentieth century. Ramanujan rarely wrote the types of proof that appear in conventional maths papers. Instead, he filled entire notebooks with formulae that he believed came from a goddess who appeared in his dreams. His work has continued to inspire new research long after he died, aged 32, in 1920.
“Eventually, humans will be obsolete,” says Zeilberger, who has pioneered automation in proofs and has helped confirm some of the Ramanujan Machine's conjectures. And as the complexity of AI-generated mathematics grows, mathematicians will lose track of what computers are doing and will be able to understand the calculations only in broad outline, he adds.
The Ramanujan Machine is designed to generate new ways of calculating the digits of important mathematical constants, such as π or e, many of which are irrational, meaning they have an infinite number of non-repeating decimals.
The AI starts with well-known formulae to calculate the digits — the first few thousand digits of π, for example. From those, the algorithm tries to predict a new formula that does the same calculation just as well. The process produces a good guess called a conjecture — it is then up to human mathematicians to prove that the formula can correctly calculate the whole number.
The team began to make the conjectures public on the project’s website in 2019, and researchers have since proved several of them correct. But some remain open questions, including one on Apery’s constant, a number that has important applications in physics. “The last result, the most exciting one, no one knows how to prove,” says physicist Ido Kaminer, who leads the project at the Technion — Israel Institute of Technology in Haifa. The automated creation of conjectures could point mathematicians towards connections between branches of maths that people did not know existed, he adds.
The Ramanujan Machine currently has limited applications: so far, the algorithms can generate only formulae of a particular type, called continued fractions. These express a number as an infinite sequence of fractions nested in each other’s denominators.
The project — described in Nature on 3 February — is named after Srinivasa Ramanujan, an Indian mathematician who was active in the early twentieth century. Ramanujan rarely wrote the types of proof that appear in conventional maths papers. Instead, he filled entire notebooks with formulae that he believed came from a goddess who appeared in his dreams. His work has continued to inspire new research long after he died, aged 32, in 1920.
“Eventually, humans will be obsolete,” says Zeilberger, who has pioneered automation in proofs and has helped confirm some of the Ramanujan Machine's conjectures. And as the complexity of AI-generated mathematics grows, mathematicians will lose track of what computers are doing and will be able to understand the calculations only in broad outline, he adds.
GPT-3 can do humour (at least taking the mickey out of scientists):
The writer and programmer Gwern Branwen got the AI language model GPT-3 to write satirical dictionary definitions of science and academia by prompting it with eight examples. Here are some of the computers definitions:
The writer and programmer Gwern Branwen got the AI language model GPT-3 to write satirical dictionary definitions of science and academia by prompting it with eight examples. Here are some of the computers definitions:
The Literature [noun]
A name given to other people’s published papers, referred to by scientists without actually reading them.
Scientist [noun]
A person who has a laboratory, access to large sums of money, graduate students, or all three.
Track Record [noun]
When scientists use this term they refer to the research done by someone else (usually a student) in order to avoid having to do research.
Faculty [noun]
Used loosely by scientists to mean any group of people with advanced degrees. Typically used when you have done something stupid and want to inform others that it wasn’t you who did it, but rather those other crazy people over there who won’t put their titles after their names.
Clinical research [noun]
Research conducted on humans, e.g. clinical trials and epidemiological studies. Researchers do not like this kind of research because humans are unresponsive and unreliable.
Speculation [noun]
One of Newton’s Laws of Science, also known as “Newton’s Law of Speculation”, which states that “Speculation breeds Confusion”, an effect well illustrated by how theoretical particle and string theorists debate scientific issues with each other. Confused? You should be. After all, you’re dealing with a bunch of theoretical physicists.
Skynet [noun]
This ersatz god, in the form of an artificial general intelligence or AGI, is what some scientists hope to build to make themselves more precious than their subjects, to achieve their goal of making all the subjects equal to themselves; in other words, to put all the subjects into a bigger tin box to make themselves feel more important.
Scientific method [noun]
A set of steps in which a scientist (or “researcher”) would probably not follow in practice: planning of research, including definition of problems, hypothesis formulation, selection and design of experiment(s), collection and entry of data, execution of experiments, data analysis and interpretation, preparation of final report, implementation of suggestions. In actual scientific practice the most important part of the scientific method is to observe and collect data without a plan or preconceived expectations about what is causing what. It also helps to be unbiased in selecting the data to include in the study and to pick experiments at random. Additional steps include writing a paper, a process that usually involves the most violations of the scientific method.
A name given to other people’s published papers, referred to by scientists without actually reading them.
Scientist [noun]
A person who has a laboratory, access to large sums of money, graduate students, or all three.
Track Record [noun]
When scientists use this term they refer to the research done by someone else (usually a student) in order to avoid having to do research.
Faculty [noun]
Used loosely by scientists to mean any group of people with advanced degrees. Typically used when you have done something stupid and want to inform others that it wasn’t you who did it, but rather those other crazy people over there who won’t put their titles after their names.
Clinical research [noun]
Research conducted on humans, e.g. clinical trials and epidemiological studies. Researchers do not like this kind of research because humans are unresponsive and unreliable.
Speculation [noun]
One of Newton’s Laws of Science, also known as “Newton’s Law of Speculation”, which states that “Speculation breeds Confusion”, an effect well illustrated by how theoretical particle and string theorists debate scientific issues with each other. Confused? You should be. After all, you’re dealing with a bunch of theoretical physicists.
Skynet [noun]
This ersatz god, in the form of an artificial general intelligence or AGI, is what some scientists hope to build to make themselves more precious than their subjects, to achieve their goal of making all the subjects equal to themselves; in other words, to put all the subjects into a bigger tin box to make themselves feel more important.
Scientific method [noun]
A set of steps in which a scientist (or “researcher”) would probably not follow in practice: planning of research, including definition of problems, hypothesis formulation, selection and design of experiment(s), collection and entry of data, execution of experiments, data analysis and interpretation, preparation of final report, implementation of suggestions. In actual scientific practice the most important part of the scientific method is to observe and collect data without a plan or preconceived expectations about what is causing what. It also helps to be unbiased in selecting the data to include in the study and to pick experiments at random. Additional steps include writing a paper, a process that usually involves the most violations of the scientific method.
And now AI can do debates
Not quite up to human expert, but not bad:
Evaluation of Project Debater.a, Comparison to baseline systems. Bars denote the average score, where 5 denotes ‘Strongly Agree’, and 1 ‘Strongly Disagree’ with the statement ‘This speech is a good opening speech for supporting the topic’. Striped bars indicate systems in which the speeches were generated by a human or relied on manually curated arguments. b, Evaluation of the final system. ‘Project Debater’ depicts the results when S1 and S3 are generated by Project Debater. In ‘Mixed Debater Control’, the third speech was an S3 generated by Project Debater but for a different motion.
How it works:
Here we present Project Debater, an autonomous debating system that can engage in a competitive debate with humans.
Conversely, ‘composite AI’ tasks—namely, tasks associated with broader human cognitive activities, which require the simultaneous application of multiple skills—are less frequently tackled by the AI community. Here, we break down such a composite task into a collection of tangible narrow tasks and develop corresponding solutions for each. Our results demonstrate that a system that properly orchestrates such an arsenal of components can meaningfully engage in a complex human activity, one which we presume is not readily amenable to a single end-to-end solution.
Is anyone here an AI? You have to tell me if you are an AI!Conversely, ‘composite AI’ tasks—namely, tasks associated with broader human cognitive activities, which require the simultaneous application of multiple skills—are less frequently tackled by the AI community. Here, we break down such a composite task into a collection of tangible narrow tasks and develop corresponding solutions for each. Our results demonstrate that a system that properly orchestrates such an arsenal of components can meaningfully engage in a complex human activity, one which we presume is not readily amenable to a single end-to-end solution.
Not quite up to human expert, but not bad:
Evaluation of Project Debater.a, Comparison to baseline systems. Bars denote the average score, where 5 denotes ‘Strongly Agree’, and 1 ‘Strongly Disagree’ with the statement ‘This speech is a good opening speech for supporting the topic’. Striped bars indicate systems in which the speeches were generated by a human or relied on manually curated arguments. b, Evaluation of the final system. ‘Project Debater’ depicts the results when S1 and S3 are generated by Project Debater. In ‘Mixed Debater Control’, the third speech was an S3 generated by Project Debater but for a different motion.
How it works:
Spoiler Interesting list of debate topics, totally would fit here :
Last edited:
Hygro
soundcloud.com/hygro/
Until the AI has an intuitive grasp of the memescape metagame we can defeat skynet.
Until it realizes that the only game-winning move is to All Our Bases Belong to It.
innonimatu
the resident Cassandra
- Joined
- Dec 4, 2006
- Messages
- 15,099
People don't understand, that intelligence itself is a good part iself also "just computation".
While driving down a road, you're computing your trajectory, if you should de- or increase your speed, taking the weather conditions into account, making adjustments based on that, etc.
It's necessary for any type of real intelligence, but not sufficient (also addressing the link from innonimatu). But AI... that can be simply math.
What else is necessary for real intelligence... that's another question.
While driving down a road, you're computing your trajectory, if you should de- or increase your speed, taking the weather conditions into account, making adjustments based on that, etc.
It's necessary for any type of real intelligence, but not sufficient (also addressing the link from innonimatu). But AI... that can be simply math.
What else is necessary for real intelligence... that's another question.
We have 3 components: calculation, a sense of self, and a re-writing of history declaring that we 'meant' to do the thing that was calculated to do.
Similar threads
