Models of our mind and communities
Google’s AI Guru Wants Computers to Think More Like Brains | WIRED
This is from an interview with Geoffrey Hinton who — to paraphrase Peter Medawar’s comments about Jim Watson — has something to be clever about. The article is worth reading in full, but here are a few snippets.
Now if you send in a paper that has a radically new idea, there’s no chance in hell it will get accepted, because it’s going to get some junior reviewer who doesn’t understand it. Or it’s going to get a senior reviewer who’s trying to review too many papers and doesn’t understand it first time round and assumes it must be nonsense. Anything that makes the brain hurt is not going to get accepted. And I think that’s really bad…
What we should be going for, particularly in the basic science conferences, is radically new ideas. Because we know a radically new idea in the long run is going to be much more influential than a tiny improvement. That’s I think the main downside of the fact that we’ve got this inversion now, where you’ve got a few senior guys and a gazillion young guys.
I would make a few comments:
- First the history of neural nets is long: even people like me had heard about them in the late 1980s. The history of ideas is often like that.
- The academy is being sidetracked into thinking it should innovate or develop ideas that whilst important are not revolutionary. Failure should be the norm, rather than the continued treadmill of grant income and papers.
- Scale and genuine discovery — for functioning of peer groups — seldom go together.
- Whilst most of the really good ideas are still out there, it is possible to create structures that stop people looking for them.
- Hinton makes a very important point in the article with broad relevance. He argues that you cannot judge (or restrict the use of) AI on the basis of whether or not it can justify its behaviour in terms of rules or logic — you have to judge it on it ability to work, in general. This is the same standard we apply to humans, or at least we did, until we thought it wise or expedient to create the fiction that much of human decision making is capable of conscious scrutiny. This applies to medicine, to the extent that clinical reasoning is often a fiction that masters like to tell novices about. Just-so stories, to torment the young with. And elsewhere in the academy for the outlandish claims that are made for changing human behaviour by signing up for online (“human remains”)courses (TIJABP).
All has been said before, I know, but no apology will be forthcoming.
Molecular biology still has its idealists, but their ponytails are greying.
Nice turn of phrase in a Nature book review by Nathaniel Comfort, of ‘Gene Jockeys: Life Science and the Rise of Biotech Enterprise,’ by Nicolas Rasmussen. I haven’t read the book but the review contains a telling phrase:
Molecular biology still has its idealists, but their ponytails are greying. Today’s graduate students want training in economics, marketing and management, even MBAs. You can now go into biology for the money.
I have mainly researched in two domains; molecular biology and molecular genetics; and for a shorter period of time (and still, to a limited degree) in using computers to help us learn and practise. Comfort draws some parallels and contrasts, between these two domains of science and engineering, writing:
The contrasts between computing and biotech are as interesting as the continuities. Unlike the nerds of Silicon Valley, who started up their companies in garages, the bio-geeks started theirs in well-stocked, mainly government-funded university labs. But, as in IT, the dream became having your little company bought by one of the big corporations. By the 1990s, both industries had become less freewheeling and curiosity-driven, and more privatized and gold-directed.
One of the differences is brought out in a quote from Stewart Brand:
“I think that hackers–dedicated, innovative, irreverent computer programmers–are the most interesting and effective body of intellectuals since the framers of the US constitution. … No other group that I know of has set out to liberate a technology and succeeded. They not only did so against the active disinterest of corporate America, their success forced corporate America to adopt their style in the end….. The quietest of all the ‘60s subcultures has emerged as the most innovative and powerful”.
Molecular biology never had this widespread culture of hacking for long—it became institutionalised all too early in its history. In the long term, the hackers will have —I hope — more influence on medicine, and most of all on how we teach and learn medicine. It is still easier to set up a software business than a hardware business. If the barriers to entry are lower, we will see more novelty. I like to think the best is yet to come.