In particular, cue exactly 1 hour into:
The session was on the challenges of globalisation, esp re: corporate culture.
Lew Platt talked about internal job boards and how HP worked as one
family with mobile parts
New head of ABB talked about how you don't want any HQ staff, but you
do want scale in purchasing airtravel, etc.
Head of Toshiba said nothing, in polished English.
Chad from Dupont had some fun anecdotes on creating a DuPont
glossary, sending out weekly all-hands email praising specific
examples of intracompany knowledge sharing, and emphasized that R&D
should have very, very focused goals -- but think laterally once you
get there. Nylon was part of a silk-attacking program -- but less
than 5% today goes into apparel. Kevlar was aimed at replacing
steelbelts in tyres -- and less than 10% goes to automotive today.
Bill blithered about email and how it can be a brittle medium for
intercompany collaboration -- any shades of trial transcripts?
Then, just before question, Chad asked Bill how he recruits. The
session goes something like:
C: Do you have any unique methods or recruiting?
BillG: Well a huge percentage of our new hiring is people right out
of university - -so having personal relationships with the key
professors in the leading schools and making sure we're serving those
professors, tracking attitudes towards our company -- particularly
when there's a dip, when we're not getting as many people out of that
institution - we want to know what are we doing wrong there, what do
we have to beef up? So it's going to the strong universities, that's
been a huge benefit for us, and that works internationally.
[So, I wonder what the bribery budget is like this year... I know
some Caltech people made off well with Adam :-)]
Charles Nesen, professor at Harvard Law: "Do you see opportunity in
Open Code for MS?"
I'm not sure that's a globalisation-related question [Laughs] So I'll
give a quick answer. There's a lot of ways software can be very
extensible. And promoting the extensiblity is a key to our success --
getting the books out there, getting the tools out there. Generally
you want to architect the software so that you can change it without
having the source code -- because if they have to change the source
code you generate [long, nasal] *in*consistent versions. And you get
a problem of quality control of all those versions. So the ideal is
to modularize the SW and publish the way people can extend on top of
that. In some cases you can't get the modularization actually putting
out the source is helpful. And I think you will Microsoft and others
doing that more and more, putting out the source code so people can
play with it, and yet still maintain quality across millions of
copies, which is so critical.
[one presume's he's referring to NT educational licenses]
Berkeley prof asks BG to compare Jini to Universal Plug & Play.
... we're partnered with HP... cooperation with Toshiba... it's
probably not worth going into here why our approach is better than
some other approach. But there will be open standards...
Ed Yardeni, Chief Economist, DuetscheBank Securities: Y2K. Does Mr
Gates have a solution?
[Moderator, give Bill a break -- Lew?]
Lew: I do not forecast a lot of cataclysmic events. But a lot of
little things will go wrong. [Whew!]
Bill: Using a packaged sw solution, you just turn to the vendor, who
has the economy of scale to make an update. Overall, it's been
healthy for companies to reexamine and accelerate the trend towards
packaged software. ... Y2K will be less than the panic some people