I know a
lot of people that have been around to appreciate the evolution of Input and
Output devices through the evolution of the computer industry. I was lucky
enough to work on the 8080 controller that had a keypad and programmed using
assembly language to simulate a lift.
One of my first programs that was a challenge in my class to build using the
minimum number of CPU cycles. I believe the contest was a draw between me and a
few other classmates for the grand prize of a Cadbury's éclair.
Next came
the computers with keyboards without a mouse. A world that required a deep
knowledge of commands, shortcuts and working on a scrolling screen with
applications on several separate floppy disks.
The mouse brought a new level of luxury, you did not have to remember
the shortcut to a badly designed menu item, where the developer forgot to
underline the shortcut key; click on it and get to the sub-items and execute
right away.
I remember
programming mouse-down, mouse-up events, then later click and doubleclick
events, followed by classes that abstracted it out. All this was gone with
WYSIWYG interface and developers did not have to look under the covers. The
amount of effort over the five odd years to get mouse to be invisible- in a
sense, something you don’t think about; was incredible.
Many of my
friends consider the touch sensitive screens as the next innovation, to me, it
was just a combining of screen and mouse. The effort from going from "raw
processing" of events to standardized swipe, pinch etc. took about five
years.
Voice
I believe
with the developments in voice-to-text and text-to-voice along with computers
today, Voice is the next leap after the venerable mouse. This may be a disputed statement, and many
people have pointed out the various limitations, the errors and the inabilities
and the chunkiness of using voice. It is
true that when you think of voice as a channel of communication, there are many
things to be desired, but if we look at the voice as the next I/O channel, the
world around us changes.
Let me put
it another way: with keyboards, mouse and a combination of the three, we- the
humans learnt the skills to operate this machine we call computers. These computers have capabilities due to
software running on them. With Voice, the machines have to be taught to work
with us; Machines bear equal responsibility in the interaction. Given the state
of the technology today, that is a tall order to fulfil.
Conversation
Humans know
conversations inherently, it is a fundamental skill mastered early on in life
with varying levels of competence. Every human has vast reserves of
information, experiences and strategies to navigate a simple exchange between
two people. We are attuned to various cues even without visual feedback that we
consider in our exchange effortlessly.
We do not yet fully understand all the aspects of how we converse to be
able to model machines to mimic.
A huge body
of researchers, scientists and engineers believe that information, precision in
interpretation and algorithms will enable computers to have voice conversations
with us. Yes, I believe this is true. It is not a question of if but when this
will happen.
Wait or Act
Do we wait
for the technology to develop where machines can have conversations? Or do we
go ahead and program conversations with the technology we have at the moment? I
believe in the latter approach. We have to start now and program with the
limited technology and get going right away. This approach will enable us to
push the envelope of our conversational understanding- the mechanics beyond
content.
Starting
now, will get us on the co-evolutionary path where we will get familiar with
the system, the interface, the design, the users and build a body of knowledge
that will enable us to make hard choices when voice is as ubiquitous as the
mouse.
No comments:
Post a Comment