"We're Writing Too Much Code"
Bill Gates Speaks About Context, Abstract Code, and Software's Future at
Carnegie Mellon University
Feb 25, 2004
11:30 AM

Fifteen minutes ago I was released from a lecture at Carnegie Mellon by the
super-billionaire Bill Gates. I thought it apt that I compose this series of
notes using Microsoft Word on my office copy of Microsoft Windows XP and
send it off as an email using Microsoft Outlook. Unfortunately the
ubiquitous computing that Bill touted was all branded Microsoft.

These are primarily a set of lecture notes I am copying and expanding on. I
did not bring a notebook or notepad with me to the lecture so these are
being copied from a torn up CD sleeve I was given outside the lecture hall
containing a distribution of Knoppix (a bootable Linux CD). They might be
interesting to some of you since well Bill Gates doesn't speak at
universities that often to the "future innovators of computer science,"
Or better the future employees of Microsoft since his visit comes the day
after a Career Fair for CMU undergrads was held in the same building as Bill
's lecture.

Bill seemed most interested, as many of us are, in the possibilities
associated with wireless. In particular he spoke of experiments with FM
sideband, where signals piggyback off existing frequencies. He talked about
ad hoc networks, wi-mesh, and p2p configurations. He spoke about the
potential for this use and rural areas and areas with highly limited access.
"Low cost computing is about empowerment," said Bill in reference to his won
discussion about issues of access. He stressed the overall importance of
productivity and talked of his own interest in expanding opportunities as
part of the global expansion that has been going on the past thirty years.
These were his social agendas that were mixed in throughout the one hour
lecture with showcases of new devices and explanations of Microsoft R & D.
Primarily he was touting Microsofts philanthropic efforts, which are
numerous but pale in comparison to the company's profit margin.

He talked about p2p networks and file sharing in particular. He spoke highly
of them in fact and was in support of the potential for amazing and legal
use to further access and pervasive networking. He spoke of the need for a
system that supports the artists but creates a filetype of usage that is
across platform rather than proprietary. Bill focused much of the lecture on
ubiquitous computing being able to have your media on demand on any device
anywhere you are. He showed some new devices but spoke of course of
bandwidth limitations being the biggest hurdle in that field.

The other major hurdle stopping media from flying through the air more
regularly are the issues around trustworthy computing. Some of the systems
he proposed to push ahead trustworthy computing are obviously taken from
many of the web experiments in friend schemes and peer approval ratings
systems we have seen. He spoke of course of the internet as a democracy (do
people still believe this rhetoric?). He spoke of one idea of search engines
that would return content rated by a friend or a friend of a friend. That we
would all be willing to rate movies, media, web sites is maybe a bit far
fetched. This does relate to some smaller networks we see emerging but is
not a breakthrough. Interesting though to my own work and much of what I see
others involved in are these small communities and primarily experimental
databases. Bill spoke a lot about databases. He has yet to look past this
model, which admittedly neither have I so let's go with it. Bill exclaimed
that there is too much code. Programmers are producing low level code that
will never cooperate and that this is why there are numerous similar
projects going on that cannot share as much as they would benefit from
sharing. This does not mean he supported Linux or FreeBSD or open source
strategies with this statement since he did expand on his statement to
declare the need for interesting and contextual visual interfaces. He was
calling for what he called "abstract code." He saw this being supported by
accepted standards.

The most interesting area of the lecture was around this concept of
contextual information. He showed some UI experiments done at Microsoft
around creating contextual information: images searchable for faces,
choosing day or night photos, text searching in movies, etc. This
information could be visualized in various 3D ways on the screen. He talked
about his continued support for eventual fruition of good speech recognition
and image recognition and spoke of the triumphs so far in text recognition.
This all related to his concept for software that enables an intelligent and
quick database. An intelligent database would be able to relate and organize
what he said within this decade would be a lifetime of media for each
individual user depending on what we want and when. A scheduler based on
this system might know where we are and what device we have and how
information should be ideally distributed to you depending on your habits -
your context. Humans understand context - software as yet does not. This is
where he saw the future of software coming from, contextual software and
data management. Creating programming environments that are not about code
but are useable graphically and based on some standards as well as
interfaces to media that are across device, have some intelligence, and
learn based on users habits, and human contexts. This is reiterating a lot
of the research I had seen in SF in 2000/2001 working in a research
environment at a design firm. This was a reiteration of the promises of
ubiquitous or pervasive computing which have yet to fully come to fruition.
However, what Bill was right about is the need for existing management
systems that organize our expanding media collections according to our
desires and preferences - not just date and filetype. He spoke of cameras
that would upload images to a server and stamp them with time as well as GPS
coordinates. He demanded software that was about priorities and contexts and
somehow Microsoft is the coming that will bring this all to us, and the
world.

Bill talked a bit about the future of AI and its uses and laughed through a
description of the robotic vacuum cleaner, the only commercial device on the
market that uses AI (according to Bill and I have no data to challenge
this). I only throw this in as a member of my own collective is involved in
research specifically about a community of hackers that mod such devices.

Most interestingly, Bill seemed most excited about the connection between
mobile gadgets, WiFi ad hoc networks, and friend networks - not to mention
blogs, Wikis, etc.. He talked about these things working together to create
a comment and approval system that would then solve many issues with
trustworthy computing (as long as SMTP standards are trashed as well).
Unfortunately for me, an artist who scammed a ticket into the event, Bill
explained that all of these innovations will come from CS and EE. He said
that the two areas of research that will improve the world in this decade
are CS/EE and Biology/Biotech. He laughed about how these areas determine
the focus of other areas like law practice, etc. He also talked about the
need for cross discipline working but several times said that the real
innovations will come from students with a CS background. Perhaps Bill
should sponsor some artist's grant programs? So all of you interesting
people that I know and don't know doing amazing research and
experimentation - beware Microsoft has similar interests and they are
cutthroat. When I worked for Palm Computing two Microsoft spies were
arrested on the 3Com campus. If they are going after Palm, they will likely
be showing up at media arts events. With so much work in the media arts or
tactical media or locative media or whatever you call it relating to friend
approval schemes, GPS, mapping, location, UI experimentation, etc, we are
likely to see that invasion soon if not already. Watch your back - Bill
might steal your ideas!

Nathan Martin
# New Cell Number 412 726 2338
5622 Woodmont Street 3rd Floor
Pittsburgh, PA 15217
__________________________________________________
STUDIO for Creative Inquiry
Carnegie Mellon University
http://www.cmu.edu/studio
iEAR Studios
Rensselaer Polytechnic Institute
http://www.arts.rpi.edu
Carbon Defense League (CDL)
Hactivist Tactical Media Network
http://www.hactivist.com
__________________________________________________
Do you Re-Code!?
Re-Code! Shopping - Clip Barcodes, Not Coupons!
http://www.re-code.com