Skip to Content

AI

Rimini, Italy

I am in Rimini attending the Perada Summer School; it is about pervasive adaptation ... Feels strange to be back to "school" :-)

Winchester, England

I am in Winchester, the AlifeXI conference starts tomorrow.

An (indoor!) BBQ will take place in a few minutes. It's raining and I am looking forward to see how an indoor BBQ works ;-) that's England!

When I left this morning Milan was hot and mosquitoes would not let me sleep. It seems that I will not have these problems here.

Looking forward to see if there are any REAL viable technologies that can be used on the Internet.

Cognitive Neuromorphic Engineering

I am attending the Cognitive Neuromorphic Engineering Workshop which is organized by the University of Zürich and is being held in Capo Caccia, Sardinia.

The place is fantastic, and the talks are excellent.

Models, super models and computation

One of the first things I noticed, when I started reading about Bio-Inspired AI and neuroscience, is the notion that we cannot say that we "really" understand something unless we find a suitable "traditional" mathematical model that approximates it in a way or another.

Mathematics Is Biology's Next Microscope, Only Better; Biology Is Mathematics' Next Physics, Only Better

While reading about Neuroscience, Biology, Bio-inspired Artificial Intelligence and Evolutionary Robotics, I was struck by the fact that it seems like in each of these fields, mathematics is stretched to the point that it is no longer useful to model system complexity. It instead gets in the way of simple and elegant explanations about emergence and models that might be able to link different levels of complexity within a system.

How many neurons do I need to do this?

I finally got an answer, which of course is derived from a model (inspired by Engineering).

But at least here is an answer, which I could not get before, and I got it reading the Neural Engineering book.

Playing with vision

Here is a little illusion game that you can play online. It illustrates how vision works.

I am reading a book called "Neural Engineering". It is a framework or theory of how the brain can be modeled in terms of Engineering, with theories derived mostly from Electrical Engineering.

Reading about ...

... Bio inspired Artificial Intelligence and Bayesian Networks.

In the past few months I met researchers from Rome's CNR, Stanford and EPFL (Lausanne). They are great researchers, all involved in this Bio-Inspired quest.

I like what I am reading, I like brainstorming about it and the idea of applying these techniques to the Net/Web is stimulating ...

Yes, big opportunity ...

Completely agree with this: "This means that AdWords ads shown in search results drive Google’s earnings, but Google searches access less than 0.02% of the world’s information. Imagine what the other 99.98% is worth.

Computer technology vs. partial differential equations

I studied Engineering and graduated with honors, but I must admit that I never liked mathematical analysis, and was never too good at it. I found it as being too abstract and hard to visualize; on the other hand, I loved computer science, it was my second nature, I could naturally visualize it without efforts. My thesis work, when I graduated, was about a simulation of an automatic storage system at a car factory, which I simulated using what are now generically called cellular automata.

Bottom up vs. top down

I personally prefer a bottom up approach to Artificial Intelligence.

I have been having many conversations about this in the past. A few days ago I decided to take sometime off "normal study" and take sometime to browse the Internet querying Google for different keywords.

I stumbled upon this old article which basically summarizes some of the stuff I have been thinking about lately. I do not agree with some of it, but very interesting read nevertheless.

Neurons

We now give it for granted that neurons fire, and basically work (mostly) digitally. It is amazing, though, that they work digitally, given the fact that the incoming patterns they are supposed to work on are not digital.

There is certainly a great incentive in being digital :-)

Books

Today I received a book I ordered from Amazon. Principles of neural science is a 1,400+ pages book.

The book is heavy, so heavy that it is almost impossible to carry it in my backpack; I have no idea how many trees have been cut to print it, and how much energy has been used to ship it, but the point is: if a device based on e-ink technology allowed me to read decently on A4 format, I would be more than willing to download a PDF and store many of these "books" in the device.

Pretty happy

I managed to install the Numenta framework on my Ubuntu PC. Had to work around a few problems and had to learn some simple Linux and Python commands, but now it works. Feels like 1984 :-)

BTW: I am back to Milan after a brief stop in Cagliari. The weather is not hot, there are very few people around, and many bicycles ... seems like a nice city today!

Human "artificial" intelligence

This week I attended the AAAI (Artificial Intelligence) conference in Vancouver.

I was a bit disappointed because I did not find much Artificial Intelligence, but rather a lot of Human Intelligence applied to solve specific problems with the aid of computers. I found a lot of mathematics and logic and a top down approach to task solving, but very little talk about bottom up Artificial Intelligence.

Syndicate content