artificialintelligence #88 : Erdos Research Institute

Hello all

Welcome to artificialintelligence #88?

This is the last newsletter of the year

On Christmas day, I made a post about the Erdos Research Institute?

It has got a lot of support including from people whose views I respect like Francesco Ciriello Amita Kapoor Ruchi Bhatia Detina Zalli, Ph.D. Ay?e Mutlu Dr. Kaouter Karboub Diana Teresa Parra-Sánchez

It's about an idea that I have been building up on?in this newsletter

Below, I include the idea from the original post - but I will also follow up in this newsletter about the values behind this initiative

?I have previously shared the idea of creating an interdisciplinary research institute to learn AI. This has been a personal project with a social motivation

?I have a name for it now and also I hope to announce advisory board members in January

?We plan to call it "Erdos Institute" - on the lines of the ethos of Paul Erdos (my brain is open) https://lnkd.in/eGAnRd_w

?Here are some thoughts

?Why do we need the Erdos Institute

a) Interdisciplinary research in AI is not easy - but is needed

b) The academic publishing model is broken - its easier to publish in the open domain

c)? Many people who work commercially in ML or DL are not really solving interesting problems

but most importantly,?

d)? We want to give the fun experience of collaborating with the best minds in AI and a Back to Uni feel which you get in places like Oxford, Cambridge and MIT

?How would it work?

a) The idea is simply that you collaborate with members to write and publish a paper in a topic you want to learn/ research

b) You publish openly (arxiv.org)

c) We provide templates and tools to publish

d) Also, advice from community and advisors

e) Initial focus in UK, Germany and India

f) The focus is mainly from the perspective of helping the members and serving the community in domains where I have a good network?

g) Domain areas: Artificial Intelligence in Engineering sciences, Life sciences, Agtech, AI and Mathematics (we will announce a list of themes)

But I am open to suggestions

h) We will have an initial pilot phase till we can prove the model?

?What is the business model?

I want to make it educational, open, social but also a self sustaining centre of learning

So, the idea is that there are three membership levels (after the free tier)

1) A free tier?

2) A contributor membership level to be part of the community ie to collaborate, contribute etc

3) A creator level which is typically when you write a paper - you may revert back to a contributor level after the publishing if you wish

4) A scholarship tier?

?I dont have all the answers

and I am funding it myself

I see it as a long term project - which should hopefully make a difference

?The most important thing to aspire for is the feeling of a true place of learning

In many ways, it's the same I try to do in my teaching i.e. genuinely create an ecosystem of knowledge sharing and make a difference to the community

I dont have all the answers - but if you are interested or have some ideas, please message me?

Values - Is it a start-up?

Is this a start-up?

No

Not by the traditional definition of the term start-up because I want to take a long term view.

I think people are becoming more aware of some of the contradictions in start-up funding today - and also are becoming more conscious of making a difference - even as FAANG layoffs will continue in 2023.

So, a startup model, in my view, takes a short-term view.

A better model is the German Mittelstand?companies which are based on generational continuity, Long-term focus, Independence, Investment into the workforce, customer focus, social responsibility and strong regional ties. If we think of the oldest companies in the world and the oldest universities in the world, they have all adopted a much more long term perspective.

I think education is really not a business in the VC sense of the word

If you consider the ancient Indian institutions like Takshila and Nalanda, they had a method of teaching which was strongly based on discussion. The same idea is reflected by Noam Chomsky (On being educated)

"Its not important what you cover in the class, its important what you discover in the class"

I spend a lot of time working with my students at the #universityofoxford on exactly this subject. Its impossible to 'teach' a subject like #AI in any length of time. Great teachers like Gilbert Strang also take a more organic approach to teaching

Hence, I don't see this venture as a traditional company - more of an educational institution with social purpose and driven by community.

For example, I think you could create a new form of educational institution based on Kevin Kelly's 1000 true fans principles.

If you have any comments or wish to participate, please message me

Happy holidays and best wishes for 2023!

Image source:

A children's book I had on Erdos

https://lnkd.in/earwGHmy