1. Given what we know about the PrinciplesOfGrok...
2. And given what we know about how humans parse data in the physical world...

There are still a couple of cognitive hurdles to overcome.

• Because we find it difficult to parse too much data being presented to us all at once... Look back to the treemap example.
• AND Because we find it difficult to synthesize information that is presented to us linearly over time... Make a left at the first light, then veer right into the small alleyway, after three buildings and a lamp post, turn left and then right immediately. When you can see the tip of the Chrysler building, head straight for one more block and I'll be waiting for you on the third door on the right.

What can we do so that people aren't bombarded with all of the information all at once WITHOUT resorting to chunking down the information over time into linear, disconnected presentations of data that are difficult for our brains to synthesize and pull together into a single coherent picture?

What can computers do that paper can't?...Animate.

Print visualizations have traditionally had a leg up on any digital presentation of data, simply from the practical standpoint that things printed on paper can be of much higher quality than anything you can do on a computer screen.

However, comparing static screen visualizations with paper graphics is an unfair comparison and doesn't take into account the computer's greatest advantage over paper: interactivity.

The ability to interact with digital data over time is what gives digital stores of data a leg up over paper stores. That and the ability to be in multiple places at the same time.

Digital data is alive and can grow and change whereas paper data are snapshots of information in time.

However, thus far, we we haven't fully exploited the organic nature of digital data to it's full potential. You might even go as far to say that what we've done with interactivty is the exact of opposite of

Innovations in Information technology have most been in the realm of lowering the cost (both in terms of user effort and actual dollars) of creating, changing, storing and retrieving data.

The net effect of such has been a massive explosion in the amount of data out in the world...not necessarily always with a positive correlation to the quality of data in the world...and possibly with a direct negative correlation with our ability to cope with and understand all of this data.

Altogether too much noise and not enough signal, that in toto, simply makes more noise.

The transition from Computer-centered design to Human-centered design

Until now, software design has been as much about what computers CAN'T do as it has been about what computers CAN do.

• [Brainstorm examples with Katie, Sheila and Lisa.]

But as computers increasingly CAN do more, we've reached a critical mass of capability which in turn prompts us to to re-examine some basic assumptions about what makes for good information design.

Specifically, we need to re-assess how information is presented to users and the mechanisms available to users for interacting with that information.

Move away from the Search paradigm of blind targeted retrieval and more towards contextual, top-down navigation.

Move away from over-simplified displays of homogeneous, structured data and more towards textured and nuanced displays of hterogeneous data in all stages of structure.

And most importantly, we need to move away from the notion of entirely replacing one static view with another, and more towards animated overlays of semi-transparent layers of data one on top of another.

Apple is already doing some of this with Dashboard widgets. Expose and Dock animations. The use of transparencies and animation keeps you grounded in a statci context while change is taking place. Apple has always done this with their menus: Never toggling or replacing menu items based on context. Never hiding or showing menu items based on context. Instead, "irrelevant" menu items are always greyed out, never removed. [See Apple menus case study.]

This approach needs to be expanded and applied in more areas of design.

Some proposals

Search: Grey out what's about to disappear. If you put the data on semantically meaninful plots of real estate, then there's semantics to what disappears too. All the stuff disappeared from the Kitchen. All the stuff disappeared from Mondays.

Don't throw out all the data at once, build it up slowly over time. Let's re-examine the Urgency versus Priority case.

We can start out with a fairly conventional, alphabetically ordered list of tasks.

• 0_Tasks_in_alphabetical_order.png:
  

Alphabetical lists aren't a very helpful way to view tasks. For tasks, what you're mostly interested in is what order you should be doing those tasks. So let's order the tasks by priority.

• 1_Order_by_Priority.png:
  

But task management isn't that simple, if it were, more people would use electonic Task lists. For one thing, not all priorities are created equal. Study for mid-terms is a lot more of a priority than Call the plumber. But Call the plumber, Get in on Nano deal and Get more aspirin are all pretty close in relative priority. Understanding the relative priorities of your tasks gives you a sense of how to apportion the finite amount of time you have with respect to your seemingly infinite list of tasks.

• 2_Show_Relative_Priority.png:
  

Then there's this notion of Urgency. Should you really be studying for mid-terms when you're toilet is overflowing and you need to go to the bathroom?

• 3_Overlay_Urgency.png:
  

Finally, doesn't the relative sizes of these tasks matter? Even though Study for mid-terms isn't particularly Urgent, shouldn't you give it some extra lead time? Paying Billy back is low priority, but would actually take even longer than studying for mid-terms because you'd actually have to go find a job and endure a period of deprivation bordering on outright penury in order to pay back what you owe.

• 4_Overlay_Task_Size.png:
  

Now, you have a more nuanced, accurate notion of how to go about "prioritizing" and "ordering" your tasks.

• 5_What_Order.png:
  

In truth, what we're trying to capture here is an approximate, but tangible visualization of your intuitive, "all in your head" understanding of your tasks. It is precisly because most software is still too dumb and too literal to capture and present these nuances that many people don't feel it's worth the effort to dump the contents of their head into a data management system.

We don't presume that we can truly capture everything that's going on in your head since that would presume that you could actually articulate to the system everything that was going on in your head. We do feel however, that we can do better than what's out there today and that even if we could only capture 40-60% of the myriad factors that weigh into the decisions you make, the benefit of having a tangible, visualization of that information would make it worth your while to interact with the system.

1. The ability to reflect upon information you generally store in your head
2. The ability to share and communicate information you generally store in your head