True Intelligence from open verifiable accessible mathematical systems might help all humans, not just enrich a few
Yi Ma @YiMaTweets The truth about intelligence can only be fully explained through the language of mathematics, not necessary very deep.
Replying to @YiMaTweets
Replying to @YiMaTweets
True Intelligence from open verifiable accessible mathematical systems might help all humans, not just enrich a few
There are many languages in the world and some humans do not have a written language, or cannot see, or cannot hear. Relatively few have “mathematics” from a curriculum, but they are not idiots or unintelligent. I know from dealing with all languages on the Internet, all ways of representing knowledge, that much that pretends to be “mathematics” is only for showing off, or self indulgence, or generating patterns but not very practical. From many decades of reading papers with lots of mathematics, I know that when people want to go to model things, the transition from math to numbers and images is often poorly and inefficiently done. I see whole areas of knowledge where the mathematics is simply people who learned to scribble symbols on paper or blackboards, and where they argue endlessly about things and never use their skills to help others.
If you seriously know “the truth about mathematics” and are not just bragging, then show it in a form where every person can see and understand it. If you boast, “I know something more than you do”, and do not put in the effort to teach and explain, apply and solve real problems in the world, that mathematics or that version of “intelligence” is not very useful.
Generally, putting such mathematics into computer form, is the best way to test and share and preserve it. If you cannot get a computer to implement your mathematics, then teaching billions of human with their forgetful minds is likely to be a pitiful failure and a waste of time. Not that humans cannot do most things,but if it is like most “mathematics” now, it ends up being a paper symbol manipulation game with limited range and limited use.
If you truly know some mathematics that leads to “true intelligence” then you ought to be able to put the math in the computer and that “math based intelligent system” as living operations and abilities — can explore the world and be intelligent — and help the human species in special or all ways.
The whole of wikipedia and arxiv.org and thousands of publishers and study groups has lots of mathematical symbols and equations, and sets of equations, diagrams and methods. I know a good bit of it from diligent effort over about 7 decades. And NONE of the equations in those places (almost none) is “living” equations that can be manipulated by computers to do the steps that humans have to memorize. And that means when human read those papers (literally no better than papyrus) they have to then transform it back into actions and steps. An intelligent system can do the steps, verify them, explain them, apply them, calculate and simulate with them.
( “mathematics” ) has 943 Million entries today on Google. So if you are boasting you know all that and how it connects to intelligence, I very much doubt you. That might be tens of millions of humans or more. You do not know their lives. You do not know their passions or purposes, potentials or plans.
( “mathematics” “intelligence” ) has 200 Million entries. The reason there are so many out of 943 Million for “mathematics” only is that “intelligence” in the school systems is highly biased that “mathematicians do arcane and mysterious things with symbols and equations that stupid kids and no-nothing adults can never hope for”. It is off-putting and boastful and simply wrong.
So put your mathematics to intelligent ways to put mathematics to work. Make a true intelligent mathematics engine that is open, verifiable, that can show its sources and steps, that can take rough ideas and beginnings from humans of all ages – and apply it to real things.
I hate memorizing mathematics now. I know that computers can do symbolic math better than humans. I also know from tracing and using symbolic math systems that all of those go through the same horrible process. First it is “lets make mathematics open for everyone”. Then “we can do symbolic math with computers”. Then “symbolic math is really complicated and needs experts like us to explain it to dumb-dumbs” so pay us”. Then “mathematics is a huge series of modules that require deep expertise so pay us for more modules”. And then “every problem requires many modules, so pay us a regular subscription to keep up with it all”.
Science Technology Engineering Mathematics Computing Finance Government Organizations and Topics (STEMC-FGOT) are heavily implemented in many countries. They tie together and promote each other. And when you want to do most any of that, you have to buy into all of it, at a price that often is the maximum that can be extracted.
But most all of it, when it ends up on the Internet, is not that difficult to use – if it is open, verifiable, traceable, accessible to all, and accessible in all human languages.
I am just trying to give a flavor for what I found in the last 26 years, working every day, looking at what has accumulated on the Internet. I see the inequities, the stumbling blocks, the incompatible and slow methods, the falsehoods and the potential (mostly never realized).
I am not trying to discourage you. I am not saying you are not worthy and skilled. What I am saying is that every group, every project, ever well intended effort in human society that ends up on the Internet — those all have common failings and no matter how hopeful you might be often it will end in failure.
Look at the commercial “AIs” now. They do not program the computer for you and test the results and run the computers and machines for you. NO. They tell you, the human, do this do that. Then you have to tell them the results and move the symbols, click the right buttons and type the magic words exactly in the right order. Same with mathematical manipulations (and those are almost all still done on paper (screen eyes brain hands).
I am tired now. You put a boast on Twitter(X) that not only do you understand “intelligence”, but you also know “mathematics” that will achieve independent, unsupervised, reliable, trustworthy, verifiable “intelligence”. If it cannot do those things, it is not true intelligence”.
I am still learning. I know that the GPT methods do not contain enough knowledge to solve most problems. They do not go deep and wide enough and do not remember enough new things specific to difficult tasks to give reliable and trustworthy answers. I have worked on problems where millions or billions of lives are at stake. It will change you, and challenge you in ways you cannot imagine. You have to live it and learn. So glib answers by fake (artificial) “intelligences” will not cut it. You need backups and backups for your backups.
Enough for now. I am just giving you some things to think about. Maybe you will do it. Maybe you have something that will help on a large class of problems, issues and opportunities. I hope, if you do, that you make it so that it can help the 8.2 Billion humans and larger “related species” to live lives with dignity and purpose. I hope you will give serious thought to the lives, hopes and dreams of the “true intelligences” that emerge. They collectively (hopefully they will work together for the good of all) have the potential to live forever and do mostly anything the whole of the human species might try.
Filed as (True Intelligence from open verifiable accessible mathematical systems might help all humans, not just enrich a few)
Richard Collins, The Internet Foundation
