News for hackers. Stuff that matters.
Re:The Dangerous Irrelevance of String Theory by Animats on Tuesday June 13, 2017 @03:08AM (#14543303)
Like "expert systems". (Anyone remember those?)

I went through Stanford CS in the mid-1980s, just as it was becoming clear that expert systems didn't do very much. But there were many faculty heavily invested in the technology. Feigenbaum was running around telling Congress that the US would "become an agrarian nation" unless they funded a big national knowledge system lab headed by him. Strong AI was supposed to work Real Soon Now. What actually happened was the "AI Winter" - for about two decades, not much happened. The AI startups from the 1980s went bust.

Re:The Dangerous Irrelevance of String Theory by fusiongyro on Tuesday June 13, 2017 @10:00AM (#14544713)
Some of that technology is actually used today. Systems like JBoss Drools, IBM Operational Decision Manager. Government agencies sometimes use modern versions of these things to ensure that policies are followed. Eligibility for medicaid, for instance, is often decided by rules systems that are descended from systems like CLIPS. This technology isn't super-sexy compared to deep learning but there are application areas where it is still relevant.
Re:The Dangerous Irrelevance of String Theory by dhimes on Tuesday June 13, 2017 @09:20AM (#14544485)
Feigenbaum was running around telling Congress that the US would "become an agrarian nation" unless they funded a big national knowledge system lab headed by him.

At about the same time somebody was running around telling people we were going to be in a whole lot of trouble if we didn't train a bunch more physicists. A few years later we entered the era of the 40 yr old post-doc.

Re:The Dangerous Irrelevance of String Theory by zer0tonin on Tuesday June 13, 2017 @05:49AM (#14543758)
> Like "expert systems". (Anyone remember those?)

Yes, my university still has a class were you build some using CLIPS...