New Family of Reliquum discovered.

Making mistakes is not easy. It requires that you tried at least. We’re talking active contributions to the field of failures. Now, we have uncovered a new family of such failures (or Reliquum). It seems that they have been around for a while, but nobody had described them scientifically, hence they are now officially “discovered”: Premature causations.

Making inferences is what we do all day, we need those to help us make decisions. But our brain is not very good at identifying the pitfalls of making inferences, resulting in sometimes disastrous consequences. It is a difficult concept to explain, that correlation and causation are two different things.

These statistics make a great job at explaining – visually – why just because two things happen at the same time, they are not necessary directly correlated. They will show you how to create your own correlation!

Who Got It Wrong First?

We are always debating on who got it right first. To whom the paternity/maternity of an idea should go to. It goes without saying (and is too often not said) that a fair amount of garbage, mistakes would have been produced in the process of producing said idea, and well forgotten.
This idea resurfaced with this nice article reviewing Gavin Menzies’ book on Leonardo Da Vinci about how

1. he might have gotten his ideas from the Chinese. And I’m thinking “Didn’t everyone speak Chinese in 15th Century Italy ?”
2. He made some errors. And I’m thinking: “What! Even geniuses are given flax for making mistakes?”

So, in the light of the controversial book, I thought I should unveil some other overlooked scientists / researchers / idea poppers. These underdogs might be unknown to most people but their ideas live, because someone who speaks the right language took up their idea…

Bose / Einstein

Buang / Wallace (and Darwin)

Constantine Rafinesque / Darwin

This leads to the term of the day: “Rafinesquian”, to mean “rejected for having recognized too many genera and species on the basis of very inadequate characters, the term being one of reproach”.

We might not be able to predict the future after all

We had hypothesised that the future of science might be found in present-day rejected manuscripts. However, it turns out now that any prediction might be worthless. This is the premise of a new book by Dan Gardner: “Future Babble: Why Expert Predictions Are Next to Worthless, and You Can Do Better”.

I have not read the book yet, only the review by world-renowned wrongologist Kathryn Schultz, whose book “Being Wrong: Adventures at the Margins of Error” I also haven’t read yet either.

There’s a lot of catching up to do if you don’t want to miss the future. I’ll start yesterday….

Failomics talk: the review

The talk at Biopolis on September 11, 2010 was well attended. I was hoping to post the video of the event, but it seems like there was a technical failure on the camera side. Never mind, there are other means of revisiting the talk. Today, I bring you some reviews of the talk.

Questions

After the talk, I was assaulted, mostly by students, demanding, and rightly so, some answers to their very pressing questions. Here are some extracts, complete with my answers:

“I enjoyed your talk, but I thought it wasn’t complete enough. I wish you could have explained a bit more, and I was even hoping I’d learn how to avoid failure altogether”. That one had me stumped. A brief discussion followed, where my interlocutor remained unconvinced that you can only produce new results by making mistakes along the way.

“I would like to know how you generated the data to produce the figure showing the generation of failures in the lab with time”. Of course, this student was referring to this animation. When I replied that I had walked into the lab and collected information that I then transformed into the animation, he remained skeptical, adding “I don’t think you could call those fractals”. I accepted graciously to remove the term. He remained unconvinced and claimed that whatever method I was using, it had to be “reproducible”. To which I replied that, like in history, you could not follow the exact same set of mistakes twice.

“Do you ever work with other types of failures?” [other than those from biological sciences?]. To this I answered that those were particularly nice to work with because the methodology came with them, but the results should apply to all failures of course.

Overheard: “I thought it might be contemporary art, when she told us she’d been growing her own failures…”. I would say: do not believe everything you hear.

Reviews

From the Prime Minister’s Office:

An email asking me to do a “factual check” on their review. This was done, and the resulting article will be in next month’s Challenge Magazine. The review insists on the “large amounts of failure required for a just a few drops of success” and turning failures into success stories (Paul Lauterbur’s Noble Prize for MRI story).

From a Japanese website

A review published here. I wish I could understand Japanese though.

On Genewired

A review by Dr Erwin Chan (SMU). Dr Chan would like to see the idea refined to get more Singaporeans to see failure as part of the thought-process.

I’d say… we’re getting there. And to finish, the only bit of video available from the event so far… yeah, technology also can fail.