From: Neil H. (neuronexmachina@gmail.com)
Date: Wed Aug 09 2006 - 13:16:05 MDT
As far as AI goes, I'd say that some of the more recent exciting
developments have been in the following areas:
* Latent Dirichlet Allocation
* spectral clustering
* boosting
* various Bayesian/graphical models
* Markov random networks
* semi-supervised learning
* scale-free networks
On 8/8/06, Philip Goetz <philgoetz@gmail.com> wrote:
>
> I'm archiving a bunch of articles from the 1990s, and remembering how
> exciting that decade was. We had a lot of new ideas about AI
> architectures, such as situated activity, hybrid architectures,
> dynamical-systems architectures; not to mention the continuing fallout
> from the 1980s over connectionism and reactive behavior. We had
> reinforcement learning, latent semantic analysis, statistical natural
> language processing, wavelets, and hidden Markov models for the first
> decent speech recognition. We had people trying to deal with action
> selection and attention in cognitive archtectures for the first time.
> We had functional MRI, brain function localization, and the decoding
> of population representations, temporal spike train coding, and
> chaotic basins of attraction in the brain. We had decent compression
> for the first time - gzip and MPEG. We had genetic algorithms,
> genetic programming, nonlinear science, self-organizing systems, "the
> edge of chaos", artificial life, virtual reality, the World Wide Web,
> Linux and open-source, developed the theory of quantum computers, had
> the first nanocomputer designs, solved the first tough problems with
> biocomputers. We also developed most of the transhumanist ideas that
> we're still playing with today.
>
> I could list a lot of things from the 1980s, also. But I can't think
> of much that was kicked off in this decade that's as exciting as any
> of the things I just mentioned from the 1990s. Support vector
> machines? Blogs? Web services? Greasemonkey? Micropayments?
> Outsourcing? That's all I can think of at the moment. Am I getting
> old? Am I out of the loop? What's going on out there? I can't even
> think of any new movements in science fiction from the 2000s.
>
> There are some exciting things in other fields - genome sequencing
> (developed in the 1990s), RNA interference, gene therapy (largely
> 1990s also), microarray protein expression analysis - but I can't
> think of much in AI/comp sci/math that excites me lately.
>
> Perhaps this is because I left the university in 1997 and went to work
> in industry - but, I never found out about any of those exciting
> things from my university classes anyway, so that explanation doesn't
> satisfy me.
>
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:57 MDT