Wednesday, May 27, 2009

Review of Beautiful Evidence

Beautiful Evidence Beautiful Evidence by Edward R. Tufte

My review

Rating: 3 of 5 stars

First a comment that I read this book because of all the buzz on sparklines a few years ago.

As for my review itself: I liked the emphasis on the power of the human vision system to process large amounts of data quickly. The focus here, then, is on high information density with as much context as possible. Tufte really likes figures right next to related text, or even within the text. He likes scales on pictures, or perhaps well-known objects for context. Also, information to convey statistical significance is also considered important, and the ability to relate relationships, too. Summary: easily understood, easily available, honest, dense information is good.

I found the diatribes on PowerPoint and sculpture pedestals interesting. I did not think he presented convincing evidence against slide presentations. He could easily have handpicked so few example sources (even the dozens he had). I saw no claim against bias except a statement that they were "unbiased" selections. He chose some people claiming that slide presentations were responsible for the space shuttle Columbia disaster, including himself (if I remember correctly). Any claims of value for slide presentations were quickly dismissed by saying that important other folks found slide presentations bad.

I find it sad that he fights so hard against misrepresented information then proceeds to use diatribe, one-sided arguments, and psychological appeals with references to Soviet oppression as ways to state his case.

I think people want information summarized in many cases. Not everyone wants or should need to read a detailed report.

So maybe the better conclusion would be, "If you have a highly-visible and expensive risk of several people dying, maybe you should err on the side of caution and be willing to spend more time and money to make sure you are right." I think that's better than "PowerPoint kills people" (paraphrased by me).

I still do find it interesting to read the arguments for real tech reports, use of standard sentences and paragraphs, and so on. Also the complaint against "pitch culture". So, even though I disagree with the extremity of his position, I think there is a lot to learn here and think about.

Side note, it seems clear that he carefully laid out each page (or pair of facing pages) throughout the book with great attention to how to final physical product would look. In that sense, this book is definitely a work of art. I don't get the impression many technical books tech presentation so seriously.

View all my reviews.

Wednesday, May 20, 2009

Tuesday, May 19, 2009

Open Source Real-Time Raytracer

A recent paper out of Stanford from Saxena and Ng mentioned that they could train computer vision from raytracer output and use that knowledge effectively in real images. They couldn't do the same with raster (OpenGL) output. It wasn't the main focus of the paper, but I still found it very interesting.

And I like fast. And I don't like supercomputers.

Therefore, I'm rather interested in the idea of high-speed raytracing. That's why I liked this blog post on open and closed source real-time raytracers.

I'd like to see a serious real-time-or-faster, open source world simulator someday. Something that could gradually add new simulation features with time. Real-time raytracing seems to be fundamental part of that (along with physics, audio, and so on).

Monday, May 18, 2009

The Answer's Already There

I've been thinking about robots learning how to act in the world around them. For any task, let's presume a program could be written to get the job done. How much effort to cover the task, including all the corner cases? Most solid software needs a lot of effort. The devil's in the details.

However, the details are all around us. Why use automated learning? If a strategy doesn't work, modify it. Automate the modification. This glosses over lots of the how question, and bootstrapping some answers into the system might speed things along. But why work out all the bugs for the system if the system can work out the bugs for itself?

I think the same issue can apply to many types of software, by the way.

The ability to sense the effects of actions is important in all this, too.

Thursday, May 7, 2009

Kindle DX for Textbooks?

So, I skimmed about Amazon's new Kindle DX. Larger than before. Supposedly good for textbooks and newspapers. Still gray scale.

I don't get it. Maybe a novel is fine in gray, but some things need color for full effect. Like picture books. Or textbooks.

Really. It's WAY easier to convey detailed information in color, and effective textbooks use that to their advantage. (My apologies in advance to those who can't see color.) Your product won't be effective for textbooks without color. That's my opinion.

Second, many students already carry laptops, often clunky ones. The Kindle might be sleeker, but expecting two devices (laptop and Kindle) seems a bother to me.

I just don't see this working. Give me a simple PDF or something (DRM'd or whatever). I'll get by. Really.