Recently Peter Norvig did an interview for Wired Magazine. It was called “The end of theory: The data deluge makes scientific method obsolete” – I know, a pretty provocative title and also a rather silly one. This is ok though, readers are attracted to bold statements and titles like that and at the end of the day their goal is to attract and keep readers.
What is not right is that they incorrectly quote their sources (in this case at least) and change the whole meaning of the thing. I have also been misquoted before and it makes for a right old mess. People email you and ask why you made such ridiculous statements and you end up having to publicly explain that you were misquoted which is what Peter Norvig did.
The write-up about this on Daniel Tunkelang’s blog is very good because it is short and comprehensive and he also adds his opinion which I thought was funny:
“I’m actually sympathetic to the view that it’s usually better to have more data than heavier theoretical machinery. But I’ve seen this view taken to an extreme so absurd as to be worthy of an April Fool’s joke–in Chris Anderson’s Wired article…”
Peter Norvig swung by Daniel’s post and posted in the comments so take a look, there are some very interesting comments from other people too, as is commonly the case on The Noisy Channel!
So…I won’t go over the Wired article I think it’s been covered adequately now (see links on TNC). I will quote some nice bits from Peter’s response because I don’t think it’s been covered enough. Read it, I think he got a huge thumbs up for setting the record straight and his lengthy write-up contains all sorts of gems for you to gather and put in your treasure chest.
These are snippets so not represented in their full context of course:
“Be gracious when you’re interviewing—let the interviewees have their say and don’t force them towards one topic. When you’ve gathered all the data, then sort out the facts.”
What the whole thing is about:
“Is it true that at your ETech presentation in March, you said, in a direct homage to George Box, “All models are wrong, and you don’t need them anyway”? Is that accurate?” (Wired)
The quote I used was “essentially all models are wrong, but some are useful”.(Peter Norvig)
what Wired wrote:
“Peter Norvig, Google’s research director, offered an update to George Box’s maxim: “All models are wrong, and increasingly you can succeed without them.”
Peter writes that: “To set the record straight: That’s a silly statement, I didn’t say it, and I disagree with it.” – phew!
I agree that writing something provocative (or wrong) can be expected of some publications but not Wired. It’s supposed to be an authoritative source. I thought so anyway. The fact that the author Chris Anderson actually said he was “going for a reaction” is…I’ll let you fill in the gap.
Chris did however as Peter notes have insight and did get some of the information right, it wasn’t all wrong. Just the misquote was really quite a big one. I guess (especially for the computing community) like saying in court “The thief was wearing a blue jumper and red jeans” when in fact it was a catsuit.
“Having more data, and more ways to process it, means that we can develop different kinds of theories and models. But that does not mean we throw out the scientific method. It is not “The End of Theory.” It it is an important change (or addition) in the methodology and set of tools that are used by science, and perhaps a change in our stereotype of scientific discovery. Today we think of a scientific breakthrough as the brilliant lone scientist, watching an apple drop or sitting in his bathtub, and suddenly having a Eureka! moment.”
I’ve written about the scientific method before (in a basic way) and for me it’s just a normal process. The fact that it gets improved and changed if just fine, this business is all about moving forwards. Why bother calling it “research” if you’re never going to try new things? These changes and discoveries are what make research exciting and sometimes even exhilarating. The foundations are so important though, they keep everything afloat.
I would love to have Eureka moments like Peter describes, but in fact my interesting discoveries are a combination of the work of many who came before me and many present now. I add my paving stone, (maybe pebble actually) to the big long road, so the next person can. It’s hard to know who to credit with discoveries because so many have a hand in it – which is an excellent thing imho.
More interesting snippets:
“Using data and statistics is not a new idea in science, and it is always done with respect to a theory” (Peter)
“I think we are all agreed that models are here to stay, and that it doesn’t make sense to talk of doing science (or computer science) without them.”
A lovely thing to find in Peter’s write-up is the Minsky-Sussman Koan (read it there). I’ve had it in all my scribble pads and at the front of my notebooks for several years. Inspiration, and a reminder. BTW I still want to take Marvin for dinner, he’s my hero (even over Batman), so if you run into to him let him know. I’m serious.
“But to be clear: the methodology still involves models. Theory has not ended, it is expanding into new forms”.
“OK, so Wired had an important insight, but missed the real story, and misquoted me.”
Hacker news also picked up on Peter’s write-up, the comments are interesting.
This is whole thing is stupid and sad. I’m always keen to pick up on things that I think are wrong, and often it’s because of journalists or bloggers trying to be sensational and provoke a reaction. This leads to people who are genuinely interested in the subject area but not experts in it taking away stupid statements and erroneous “facts”. This does not help them one bit. It only benefits the publication. That’s abusing their authority and being the starting point for an awful lot of blog posts and comments all over the web that are full of errors. These continue to perpetrate these falsities.
Wired finish with the question “What can science learn from Google” – personally I think that the entire point of the research community is to share, discuss and innovate. I don’t see Google as a separate entity in that community but as a valuable and highly experienced contributor. Google as a whole (researchers, computer scientists, coders etc…) are in science.