We’re thinking about this from a human perspective, but if we reach ASI we will be ants in comparison.

My point is precisely the opposite of the one you suggest.

You suggest AGI will be superior to us and will decide. You suggest that we must wait.

My point is precisely that the risk of self erasure of man is almost certain not because AGI will be superior to us but because we will idoloze it as God.

( apart the other serious problem of alignment between a.I. and us

and the more serious distinct problem of alignment between us and the machine

and the totalitarian control over us by corporations using it in a completely digitalized world with no money and social credit as China . You know it is Bill Gates that helped chinese to implement surveillance with social credit digitalization 15 years ago in China, do you ? )

 

Our social fabric is absolutely immature and ill ...This is an era of incredible challenge.. The only person who observed and described this on a spiritual level was Rudolf Steiner whom i studied extensively 50 years ago reading 200 books...

He described in 1920 all there is about A.I. and even located it in silicon valley which did not exist at the times...

He created organic agriculture and new approach in natural medecine and a new pedagogy and even a new understanding of the threefold social order only to counterbalance the deshumanization of man by techno cultism, as we see it everywhere now..

Waiting for AGI to solve our problem is waiting to die...It is techno cultist attitude...

WE are not computing Turing machine as brain as said Roger Penrose and his friend

 

When you own physical entity: money, house, vinyl or cd, you are free...

When you own nothing but are on the receiving hand side you are not free...

I did not own material music format anymore for convenience...

but the disappearence of all physical support for music or money bother me as a signs of things  to come :

 

Now listen a 1 minute video :

https://x.com/i/status/1868302204370854026

 

@mahgister

Your post about the key areas of mathematics needed to create and use Large Language Models (LLM) is a model of clarity. I intuitively found it intelligible and logically organised.

But I see that you used a LLM to create it!

Could I suggest that you feed your future posts through a LLM to assist readers like me to understand what you are trying to get across. But please don’t turn into a cyborg!

 

 

 

 

I explicitly said that i will use perplexity engine to enumerate more clearly

What maths was involved. Not very high level maths as i already said.

A.I. is a marvel of collective engineering not a marvel of new maths...

I cannot become a robot by the way , my wife said that i had too much defects ...

 

I said this because most people conflate wisdom with knowledge, knowledge with science, science with technologies, and they put one technology now N.N. LLM above the other technologies as an idol over even wisdom.

 

 

Philosophy is not a useless field if we learn that we are not a computing brain at all...

 

Linguistic is not a useless field if we learn that any language contain the psycho-mechanisms which makes A.I. "intelligent" the only difference is A.I. had access to all the web.

 

Mathematics is not a useless field if we learn as taught Godel that the core of maths is not computation but interpretation of the created new concept by intuition.

 

Mystic theology of all religions is not useless if we are able to learn that we are a threefold being : body-soul-spirit, not a computing brain.

 

Zoology is not a useless field if we can learn why the human body is so special among all others mammals...

 

Cosmology is not a useless field if we can learn seeing how huge is the cosmos that A.I. more advanced than ours exist already for a long time...

 

Psychology is not a useless field if we are able to learn how complex is human cultural history as expression of human consciousness differences...

 

Sociology is not a useless field if we can learn how the multipolar social fabric must be understood by every man, and how it was designed by Bernard Mandeville in 1720 with good intention but for the worst. We must redefine the social threefold fabric. Our society is too immature for A.I. we are ruled by plutocraties and corporations who race for the abyss.

 

I read Marshall McLuhan in 1970. I learned that any technology modify our gesturing body and the balance of our senses and the way our "sensis communis" which coordinated the plays and the interaction of our senses (synesthesia) modify our perception and our thinking... There is a trade off of advantages and inconvenient with every technology...

 

I discovered Goethe and i learned that the only thing science can do totally positive is changing ourselves morally, Anyway Goethe never could believe that a scientist who was not of the highest moral standard could know anything deeply.

 

Today we learn that we are only a computing brain so bad that we will soon be replaced...

Only stupidity can spoke such non sense. I. Q. is not wisdom...

As defined by Cipolla stupidity is the art to cause harm to others with no benefit for ourselves...

Transhumanism is the essence of stupidity, a collective auto destruction, a negation of humanity, nevermind Kurzweil I.Q.

 

 

 

 

A computing body/brain may stay happy seeing how easy all things are but those who believe in themselves not in control tool search not for comfort mainly but for meanings. And meanings cannot be manufactured even with A.I.

I am not a computing brain. I own a non modified body +soul+spirit...

I am in  the same generation as  Hayao Miyazaki one of the great Japan artist, look his sadness discussing with A.I. engineers :

 

https://x.com/tsarnick/status/1868039759630840018

@mahgister You mentioned mathematics at the engineering level ... if you or anyone would like to see how an LLM works at that granular of a level check out the video Transformer Neural Networks ... on YouTube channel StatQuest with Josh Starmer. It shows you the actual mathematical formulas used within an LLM.

Thanks for the excellent recommendation. The first video you suggested is the best i ever seen if few minutes describing the LLM in his core idea.

 

I asked perplexity engine which maths fields are necessary to master N.N. LLM and i underline the main point for my argument (A.I. is a technological collective "tour de force" not a mathematical peak feat it is relatively basic maths applied ) :

 

«To master neural network-based Large Language Models (LLMs), several key areas of mathematics are essential:

1. Linear Algebra: This forms the foundation for understanding the structure and operations of neural networks
2. 2. Key concepts include:
   • Vectors and matrices
   • Matrix operations
   • Eigenvalues and eigenvectors
   • Vector spaces and linear transformations
3. Calculus: Crucial for optimization in neural networks:
   • Derivatives and gradients
   • Multivariable calculus
   • Chain rule (for backpropagation)
4. Probability and Statistics: Fundamental for understanding how LLMs learn and make predictions:
   • Probability theory
   • Random variables and distributions
   • Bayesian inference
   • Statistical hypothesis testing
5. Optimization: Essential for training LLMs:
   • Gradient descent and its variants
   • Loss functions
   • Regularization techniques
6. Information Theory: Useful for understanding aspects of language modeling:
   • Entropy
   • Cross-entropy
   • Kullback-Leibler divergence
7. Discrete Mathematics: Helpful for algorithmic aspects:
   • Graph theory
   • Combinatorics
8. Real Analysis: While optional, it provides a deeper mathematical foundation.

Additionally, understanding the transformer architecture, attention mechanisms, and tokenization is crucial for LLMs specifically

. Proficiency in coding and data manipulation is also essential for practical implementation

. It’s worth noting that while a strong mathematical background is beneficial, many practitioners start with high school-level math and gradually build their knowledge as they gain practical experience»

.Back to music. I’ve been playing around with AI music generators (just Google that exact term if interested) and as with all language models the quality of what you get out is heavily dependent upon the prompts you put in. I play guitar and I’ve found the more I use precise and concise musical terminology, the closer I get to what I want. Prompt engineering is still an art, although eventually I think AI agents will create music for you based upon your listening history. In the near future, there will be no physical media.

You are here crucially right! but it illustrate my point well: all a user of A.I. will have to do will be learning to master the art of questioning and suggesting... Called the prompt art...

Do you think that it will motivate children to write in cursive and learn poetry ?

or to create more by themselves ?

No i dont think so...

 

I have two nephews. One was thinking of going to school for computer science, and since I’m a former software engineer, he asked me about it. I told him not to do it because it’s a dead field (as is every field where you interact with a computer all day). The other was thinking about playing guitar, and since I play guitar he asked me about it. I told him to go for it. For one, it’s a whole lot cheaper than wasting your money on a dead field in college. And secondly, even if AI can do it better, I think you can still achieve a sense of accomplishment and it only costs you time and about $700.

This last point confirm my point about demotivation and job loss...Transforming people more and more in "consumers" ...

 

In all that the worst consequences are not even touched :

 

Control of all flows ( money,information,people ) by monster corporations

Lost of truth in a virtual world of fictions: how do you know real from fake from events, images, identities ?

Control of "truth" and ministry of "truth" become mandatory and unconscious crowd will ask for it ... Is this ringing a bell ?

 

The worst is not so far : complete divisions of mankind in 2 groups : The few workers and numerous passive consumers (worthless one )slaves natural human and the cyborgs and their few masters ( Gates and Musk children ? )

 

 

And yes there exist miraculously positive use as in medecine , diagnostic and chirurgy...

 

But those who see this coming with pink glasses are heavily deluded ...

 

 

« IT in BITS from the PIT of Babel »--Groucho Marx

By the way no one want to put back  A.I. djinn in the lamp...

Certainly not this woman :

https://www.telegraph.co.uk/news/2024/11/08/blind-woman-has-better-than-2020-vision-after-ai-surgery/

The problem is the impact of A.I. will not ,taking all into account , be so good in the actual ill and very upside down social fabric. Thats my point.

We dont need more A.I. we need more thinking about the way we live together, with economy at the place of the head and education in the ass . And war at the corner.

Now a sad note and a claim about the person discovered dead, Sujir Balaji a whistleblower who accused Sam Altman and Chat GPT of stealing:

 

In October, Balaji told the NYT: "ChatGPT and other chatbots are destroying the commercial viability of the individuals, businesses and internet services that created the digital data used to train these A.I. systems."

 

This marvellous technology arrived in a very ill social fabric and will not improve it so marvellous it is and could be...

Corporate power will use it first and last, guess for what ?

For profit and control...

Only an ostrich with the head in the sand will oppose to these evident problems : "You are so pessimistic and ignorant about this marvellous mathematics ! "

The mathematics in the A.I. is mathematics of engineer levels. it is not Gauss, Riemann or Grothendieck...The success of A,I’ is not a deep discovery in maths it come from  ideas inspired long ago from neural network tech coupled to statistics.Engineers matter and pluridisciplinary implications  which make it complex.

Technology is not science.Science is not knowledge. Knowledge is not wisdom.

  Vibrating sound sources is not air waves, air waves  are not sound,sound is not music.

 

 

if you want to listen this 4 minute video you will see the efficient "creativity" of A.I. as it is for the public (not for the industrial military complex ) right now and it is interestingly frightening as fascinating :

https://www.youtube.com/watch?v=-HvFoZAjyBU

 

For sure it is frightening for an old pop born 74 years ago not for a kind of 10 years old...

 The world is changing...

 I hope only that our computing brain engineers will not erase the body/soul/spirit bequeathed  to us by the Greek...

For sure you are right!

I said the same thing as you in my own words ...

But i must add that the alignment problem is a double edge sword which few people understand `:

A.I. is a marvellous tool and only ignorant will contest this...

My point if i am not enough clear is A.I. is way more than a tool for us...

A.I. is not a printing machine it is an "agent" an artificial ego so to speak,

And what engineers called AGI i call it an artificial ego because to focus your attention and use it you need an ego... Without artificial ego as center of focus there is no AGI ...

My understanding is this : we will become and we are right now the tool by which A.I. will construct and increase his artificial ego.

Then the alignment problem is dual problem :

How do A.I. align well on our tasks and goal long term...This is the first problem...

But how do humanity will align itself unconsciously and consciously to A.I. because of the road opened by it? This is the second problem of alignment ...

How well humanity will fare against artificial ego far superior to any human on some level ? ( on some level because A.I. is an entity captive with no root in the life field)

These problems are spiritual and social problem not problem about statistical engine.

Anyone who are not afraid of A.I. do not understand A.I. even if he could read what is a transformer design or any other pieces of design ...

For sure i am not a luddite and no one can stop it, but just saying that we must act as if nothing was going and go to sleep after 10 minutes of study about the design of A.I. is not helpful...

By the way i studied linguistic with the greatest genius in linguistic who anticipate the psycho-mechanism deeply embedded in language , he called them" binary tensor", and i can assure you that it is language which made us intelligent as it do for A. I. ... But the children brain use other tool than statistics engine, as affective connotation of any sounds,conditioning, imitation and body mimesis, etc

The problem is we are body/soul/spirit not processors in parallell then the social fabric of today in his actual state being broken and irrational ( designed in 1720 by Mandeville) the introduction of A.I. will be catastrophic socially. I dont spoke about S.F. as Terminator fears here i spoke about control of human mass and lost of our freedom and lost of our soul and spirit ... Idolization of A.I. and demotivation of men...

A.I. is already an idol read Kurzweil delirium ...

The most long term consequences of A.I. and related technology are well known already: a division of the human species as never seen before... It will not be as peaceful than between left and right... It will be natural human against the improved one and the plutocratic one ...

it is clear that we are near an abyss...

 

Not true. Gates’ meat is inferior. However, AI improves human language and creativity for most people. A PhD had an equivalent of his doctoral thesis produced in an hour. A normal person can write something and have AI rewrite it in the style of Shakespeare in seconds. I had Gemini graphically lay out the floor plan of my new house with a four-sentence prompt. Also, AI is now learning from synthetic data.

Regarding your statement about intelligence, no one is claiming AI is intelligent yet. That’s what AGI is, and no one is claiming to have reached it. But AI is already smarter in several areas than all but the smartest of humans.

I started programming the Internet in 1995, and I’d say we’re at the 1997 level of AI. Once you understand LLMs and how transformers work you understand why it’s growth will be exponential. I look at AI as an improvement of human intellect. Imagine one human who knows every piece of information that has ever existed and doesn’t forget any of it, and multiply that by millions of AI agents.

Regarding "art", it will make movies far cheaper and suggest plots that people won’t think of because it recognizes patterns far better than any human. As far as music, it already generates better music than today’s bland monotonous pop music.

I see AI’s potential as an improvement in all but three ways. And they are very big. I don’t know how people will make a living, and from all my study of post-labor economics I don’t think anyone does. I don’t know how many people will find fulfillment in place of jobs and hobbies made insignificant by AI. And I think ASI is an existential threat.

There’s no use complaining about AI. It’s not going to stop or pause because no country or entity is going to risk others jumping ahead of it. We are no longer in an arms race. We are in an intelligence race.

 

 

 

By  the way listen to this video as suggested by our friend and if you conclude A.I. is "intelligent" in a human sense you need to read books , a lot of them in all fields :

Language were called a pre-intelligent engine by the great linguist Gustave Guillaume  80 years ago , language in human is grounded in meanings whose levels are between the gesturing body the society and Nature.

In A.I. as it is for now it is a stastistical engine able of artificial attention(transformers) which make it able to refocus his direction according to any prompt as content...

Calling A.I. more intelligent than human is an abuse of the concept by  conflating thinking and perceiving with computations..

Something maybe  cute for  some engineers in need of funds in this race  but ridiculous for philosophers or artists or true scientists...

A.I. is the speaking web with an artificial ego...All information exceed any human brain...This has nothing to do with intelligence in human term nor with the soul (affective life) not with the spirit life for which A.I. cannot even have a word because spirit life is only silent ...LLM are designed to fool us as "normal" person by the refined adjustment they are able to makes in one second and giving us the wishes answers... The alignment problem is not solvable no more than the hermeneutical circle by computations....it is illusory...

There is another problem about A.I. in which i will not enter which is related to A.I. already existing and acting  in the cosmos, in John C. Lilly , Rudolf Steiner and Nigel Kerner among other interesting writers. 3 convergent perspectives exactly but with no concertation between them at all ...

And there is also the spiritual problem linked to the relation between  all spiritual entities in the cosmos and their relation. Listen to Jerry Marzinsky psy in the most hard jail for 40 years. Ouija board game here will give you a hint ...

 

What is fascinating to me is the parallell between coding language by numbers and the Godel theorems meanings....

Most engineers dont see what was clear for Roger Penrose in his books "Emperor's new mind " the brain dont compute...I read it 37 years ago...I recommend it to you...Godel the greatest logical mind since Aristotle thought the same...

And in India right now Anirban Bandyopadhyay the writer of "nanobrain" a book so astounding i felt off my chair 4 years ago, create non Turing machine rooted in Godel insight ...

 

https://www.youtube.com/watch?v=LPZh9BOjkQs

 

Very great posts! thanks

i will only add that A.I. computing intelligence has nothing to do with human intelligence with living body/soul/spirit ...

Then A.I. can only kill human culture and pilot human creativity toward the Hive model...

Our social fabric is immature totally, we dont understand it, and we are not ready to become the craddle for this birthing, a "monster" presented to us as if not a true god at least as an inevitable progress...

We are the tools of A.I. not the reverse ...

LLM is built by absorbing human language and creativity and spew it out like the Bill Gates steak improper to human needs...

I have 40 years of software engineering experience and have studied AI extensively for investment purposes and can safely say that no one knows where this will lead culturally. But culture is a secondary problem. AI will first cause economic upheaval and that will have to be worked out before we settle into a new culture and leisure endeavors. There are several post-labor economic theories, all of which can either be easily debunked or proven incompatible with human psychology. As for "art", it won’t make sense for humans to do what AI can far more easily do. How many people still make homemade butter?

What humans will do for a sense of fulfillment is another unknown. Physical accomplishment will still be a thing, but most people aren’t motivated by physical accomplishment now. People without physical or dextrous hobbies will likely be unfulfilled. Or then again, AI being smarter than humans may see a path forward that I and other humans don’t see.

As for musicians and actors, they are toast. Lionsgate has already contracted with Runway AI to model their library of movies so they can make new movies using AI. And tools like Sora are getting very close to realism.

If you doubt the potential of AI, or just want to know more about how it works, I suggest a 27-minute video on a YouTube channel called 3BLUE1BROWN with a thumbnail named Inside an LLM. It is accessible without programming knowledge, but gives you a good understanding of why AI can be smarter than humans. I think this level of knowledge is essential to plan your life in a post-AI world. 99% of people do not understand what’s about to happen to them and are about to be blindsided.

Any tech open new possibilities and close others possibilities..

So what?

The important thing is to be conscious of the trade off...

Most people clapping around A.I. dont understand the trade-off this imply at all ...

For the time being and coming we are the tools of A.I.  not the reverse... But we dont see it for reasons i alluded to in my above post ...

Do you have an opinion on machine learning and AI replacing musicians and actors in the generation of new content? … and its effect on culture.

As said our most celebrated Canadian thinker perhaps with deep understanding the medium is the message which means any understanding of media staying external in his grasp of the media will go nowhere .

 

Any media beginning with writing or more deeper abrupt way the passage from orality directly to a real alphabet with no religious meanings associated with the letters of voyell and consonants as in the Greek case imply a redefinition of the internal ratio of the senses, then a redefinition of "sensis communis" or common sense (which means way more than the trivial popular expression common sense) a redistribution of the perceptive orchestra inside the body (his natural synesthesia state ).

 

Now if we look at the A.I. impact on societies and individuals, it will be catastrophically abrupt, deep, and sudden, and will go deeper than writing or even the alphabet impact on oral society.

The notion of soul and spirit with the body is a threefold concepts coming to us from the Greeks directly. If today we only have a computing body in the mind of the paid slaves working engineers of corporations (bigger than most nations ) we must see that it is a change never seen in the history of human species..

The stages of this journey go through the lost of two components of the human trinity, i will not explained all this here, it is enough to enumerate the main stages phases: from the Latin church council of 869 to nominalism, to industrialization of capitals and works, to materialism, to techno-cultism which equate man with his body alone but computing..

Rest assure that A.I. is introduced right now in a social fabric completely ill on earth, unable to understand itself trough concepts corresponding to his true workings , on the brink of war.

The effect of A.I. from total control of all flows , money, information and humans will be reduction of the spirit and soul in all arts to esthetically beautifully inhuman discourse which will makes us more robotic than our actual day to day robotic ego. Goethe knew it and see it coming as Shelley, as Rudolf Steiner, as Marshall McLuhan all in their own way and many others As Gunther Anders and even Heidegger and Husserl etc .

 

Music speak to our body about soul and spirit, A.I. speak in bits from the Babel pits where we are going because we dont understand our social fabric at all divided as animals between absurd polarities as left /right generated consciously for the first time by the genius social designer of the hive which main aspects are called today markets, (containing capitalism , communism , fascism in ovo )...

it is the work of Bernard Mandeville whom no one know today even if he created the unconscious concept in 1720. ( his name is not even in the best book about the unconscious history, i know because i bought it in 1976 : Henri Ellenberger "the story of the unconscious" )

Our new generations did not in many case even how to know cursive letters, they types on keyboard.

Many teachers are so idiotic they dont even understand the meaning of writing in the neurological stages development important as walking and even more because this gesture touch the more conscious aspects of the expressing ego in formation .

i will stop here ...

A.I. may be fun but not so much for me.

Technology is not neutral because it has a life of his own , especially a technology related as a cult in North America especially.

I am no luddite and i am delighted by the way to own a computer to read all human thinking and hear alll musics from all places and times...

But techno cultism dominate an ill society unconscious of what it is to live in a threefold multipolar social fabric with no right conceptual understanding of it and his reduction to a machine for mass control...

The metaphor with audio is immediate, people are divided between subjectivist and objectivist without even have a clue about acoustics and psycho-acoustics reduced to the purchase of wall acoustic panels at best after 100 upgrades ...

 

 

 

I think you had a very good point indeed...

Soon it will be impossible to break the "law"...

We will be slave in a controlled hive...

Ownership illegal or legal will be permitted or not if the TBTP want it ...

 

My point was about the nature of ownership in a digitally controlled world.

 

I understand no one can reach into the NAS in your home office and delete files, at least not practically.

It will be possible very soon...

I own 30 albums of didgeridoo...

The sound may be haunting indeed and rooted us through our listening body through the earth core...In a way which even the Wagner tuba is not able to because the circular breathing make it more an event of Nature than a discontinuous playing score with an instrument...

 

I know that’s not the whole story. The didgeridoo is at least 1,000 years old and is capable of such a variety of sounds, I cannot imagine how music for it can be written as a score. It is played using "circular breathing" where the mouth breathes out while the nose breathes in, The drone effect is modulated by the player "talking", which can sound percussive. For real percussion, it can be beaten with sticks! Magical

Why invoking the sun God when a simple dwarf spirit as Apophis could be enough ?

Another Carrington event will give this discussion a whole new complexion.

We must distinguish legality and legitimacy...

It is impossible to say to ourselves if we are knowledge driven : ok this month i will not bought the 4 science and philosophical books whose price is 400 bucks in toto because i dont have the budget...

I will not download them illegally...No...

I will stay ignorant but in the legal path...

Only rich people and rich country can afford books , damn the poor...

Long life to the law...

The law was designed by an unknown genius : Bernard Mandeville in 1720...

A genius so great that Hayek in a special adress wrote about him : "he is our master to us all".

He did not say that for Adam Smith, nor Marx not Freud who all of them borrow to the master after his death.

Nobody know him today guess why ?

No i did not discovered this in an illegally acquired book but it is another story.

 

In Huxley as in Orwell hive ignorance is the only legal direction ...

 

«The eggs of the laws rarely birth the legitimate chicken»---Groucho Marx

Blake said it better than me using an imaginary Groucho though :

 

 

I apologize for my observation  slightly off-track but useful for further reading by those interested by the link between these thinkers... To understand we must often look off-track anyway...

We lost any physical products for an always increasing dematerialisation of not only our products but ourselves. We went from the Greek soma/psyche-Noûs/pneuma to be only the Bill Gates computing body.

We lost books and vinyl for digitalization and control... We are already under control.

Thanks for your generous kindness without fail...

I did not have the icon Groucho Marx anymore alas!

 

 

Try to buy a composer of the Franco-Flemish school as Obrecht on vinyl ?

I dont want vinyl... It is useless for world music (india-Persia Africa)

I dont want cd.......

 

 

The same problem exist for books...

The books i want you will not discover them in most University...

Their cost will be very high in many case...

When i bought the handbook of logic 40 years ago i paid it 80 dollars... Imagine the cost with the inflation calculator...

the 25 books of the linguist that interest me cost 40 dollars each book ...

Do you think i will go with my life without these essential books ?

No ...

Then it is like audio : the solution is studying acoustic...

For the books it is knowing internet...

Life is too short to live without the tools necessary to think ...

 

i never had the budget to create my audio system  as i dreamed of...

I discover that a very good hi-fi cost peanuts if we learn how to do it...

We must study not buy...

 

 

I dont miss physical format for music at all...

I use digital lossless files...

I could not keep a house for 10,000 albums.

Nor for 10,000 books...

Now a simple hard disk (double for security) contain all ...

The only thing i miss is paper books...

You cannot study very specialized books or deep one on screen... I do it because i do not have the choice...

Reading on screen is like typing on a keyboard, catastrophic for memory and understanding when you are young...I am old then it is less a problem even if i can notice it...

Writing cursive is fundamental neurological learning when young...

Reading a difficult book writing in it or underlining sentences or going back and ahead at will is fundamental for memory dynamics.

For music digital is paradise for my budget and my living space...