[All errors are mine. The views for information only. These are views of Edward Tufte, not IBM].
Organized by The Hive
- Yesterday · 6:00 PM
Edward Tufte will discuss seeing, reasoning, and producing in high science and high art. His current project “The Thinking Eye” suggests self-aware strategies for improving the interface where the real world meets the human eye-brain system.
Main goal of principled deisgn cycle is not for audience but for
long time horizon. Collect, create and offer it to the world -- world
can take it or leave it.
To collect, you need vacation eyes -- thinking eyes -- fresh eyes.
Avoid stake metaphors, prejudices, be self aware, aware of prejudices and
go with an open mind.
Exprience things first hand, rather first eyes -- with your own eyes,
not photographs, not movies, but with your own eyes. That's when you see
data collection bias, sampling bias. Don't just depended on aggregated reports.
Your thinking starts are your retina. That data is transferred to brain at
20 MB/sec. Since there is ton of data coming in to the brain, brain continuously
tries to ignore and conclude! Trying to take in data and conclude often is one of
humanity's big sins.
When you see things, use all of your focus on what you're trying see and experience
He summons -- shutup and look. Talking & listening takes lot of processing power.
Next time you're in a museum, be silent. Look -- again and again.
Practice Intense seeing.
Look at real things -- as much as possible -- and not their represenation.
Tip: When you search for anything google, search in google image search instead of default doc search.
You see the work instead of words.
Tufte showed results of feynman diagrams (and couple of other things I don't remember).
feynman diagrams: http://bit.ly/10b0SPX
I tried some... It's lot more stimulating to see the results.. This will change all the assumptions
about search engine optimization:
Of the many companies I search, only Ferrari actually showed its products! http://bit.ly/108JASA
While showing results of the image search, the "space" between the images take about 24% of the
available space! Professor showed a re-rendered page -- more images or bigger imanges with same space.
Use space wisely -- show more, show bigger.
When you write or organize, focus on verbs not nouns.
Design mimics hierarchy -- see the internet proposal doc by Tim Berners-Lee.
Internet flattened the organization & society. Now you can send and get message from/to anyone!
Bigdata has been very useful and successful in aiding the development of natural sciences over
last 400 years -- physics, chemistry, astronomy, cosmology, geology, biology and more. Those
trying to use big data for social science, should have some modesty.
Just having a bigger pile won't help. Focus on excellence --
elegance takes care of herself. Long term horizon seeing is heard, really heard.
The only way to progress is by double-blind studies -- collecting data after the fact
will lead to wrong conclusions -- aka stylized facts!
It took 35 years to prove HRT causes cancer -- to prove conclusively.
For PSA, NNT (Number Needed to Treat) is 48. i.e., you've to test patients to improve
health of one person.
Big data analysis, done wrongly, can create too many false alarms. You can't discount the
cost of false alarms.
Brain is indifferent to mode of presentation.
Do "whatever it takes" to represent the data in a way to accurately understand, think, interpret and compare.
That's a message worth remembering and spreading.