managing innovation: artificial intelligence
DESCRIPTION
My team investigated closed vs. open systems of innovation through the lens of a particular technology: Artificial Intelligence. I took a pretty large risk in taking such a deep mathematical tone in the beginning, but think I did well to keep it accessible and relevant.TRANSCRIPT
WHAT IS ARTIFICIAL
INTELLIGENCE?
CONTRADICTIONS
“Infinity is infinity but infinity doesn’t adequately
contain itself.”
Russell’s Paradox of “class of all classes which are
not members of themselves
Illustrated: Barber of Seville
Each man in town either shaves himself, or goes to the
barber
The barber shaves only those who do not shave
themselves
Who shaves the barber?
GÖDEL'S INCOMPLETENESS THEOREMS
Proved that the systems of mathematical logic are
flawed.
No matter how large you make your set of axioms,
in arithmetic there will always be statements that
are true, but cannot be proven so.
Another way of saying this, for us technology
managers: No matter how much data you have,
even infinitely many data bits, you cannot prove all
true statements.
FROM ESOTERIC TO CONCRETE - DISRUPTIVE
INNOVATION IN MATHEMATICS
Alan Turing – The Halting Problem
For mathematicians, how do you know if the problem
you are working on is inherently unsolvable (Hilbert’s
Second Problem), or extraordinarily difficult (Fermat’s
Last Theorem)?
In conceiving an answer, Turing turned to something more
basic: uncomputability. What are the limits of
computation? The machine he constructed, The Turing
Machine, was the conceptual creation of what we today
call the computer.
WHY DOES THIS MATTER?
Every single one of us will have our lives inexorably
and profoundly changed over the coming decades
by AI.
It is important because it tells us what AI is NOT.
Internet ≠ TV, only better
AI ≠ human intelligence, only better (perfect memory)
Establishes a respect for the AI technology, but a
deep and abiding admiration for the natural
technology of the human mind.
FROM IDEA TO INNOVATION
Innovation = Commercialization of Ideas. AI
developed only as an idea until the hardware could
catch up. Now, with the situations somewhat
reversed, funding is pouring into AI research.
DARPA’s CALO project in 2003
Trapit and SIRI
VOICE RECOGNITION-CHALLENGES
speaker dependence,
continuity of speech,
difficulty of identifying word boundaries - as in "youth in
Asia" and "euthanasia.”
vocabulary size
Large vocabularies cause difficulties in maintaining
accuracy, but small vocabularies restrict the speaker.
APPLE
>50,000 employees and with annual revenueapproaching $100 billion grow 60% a year
Multi-focused structure in which product,function, and geography are emphasized all atonce
Better alignment between functional and divisional goals
Simplicity is key.
It is deceptively straightforward with none of thedotted-line or matrixed responsibilities popularelsewhere in the corporate world
A corporate dictator who makes every critical decision(Steve Jobs)
APPLE
A cutting-edge startup rather than the consumer-
electronics behemoth
The attention to detail, the secrecy, the constant feedback
-- into processes
Passion for innovation and an uncompromising
commitment to bringing great products to market.
Smart technology also needs to be beautiful technology
HOW APPLE WORKS
Accountability from top on down
a series of weekly meetings
never any confusion as to who is responsible for what.
The "DRI" or directly responsible individual.
Ability to move nimbly
Ability to focus on just a few things at a time
Still a startup at heart
Most notably by putting small teams on crucial projects
Do-more-with-less mentality
HOW APPLE MANAGE
Value-driven business-model innovation
Smart technology (ipad, phone)
Voice recognition is a disruptive
technology, but they apply it as a sustaining
innovation
Acquired SIRI (2010) ability to correlate data
ability to interpret meaning
If improved upon,...
LESSONS FROM APPLE
Network Innovation
In pursuit of Simplicity
Fail Wisely
Not All Innovation is Equal
Innovation Doesn't Generate Growth. Management Does
GOOGLE’S VOICE RECOGNITION
Application
Simultaneous subtitle in video
Perspective
Translation
A supercar in Knight Rider
and Green Hornet
Searching by oral inputs
GOOGLE’S CHALLENGES
Challenges
Vocabulary
Accent
Automatic skip
Translation
MICROSOFT VOICE RECOGNITION
Windows Speech Recognition
empowers users to interact with
their computers by voice.
It was designed for people who
want to significantly limit their
use of the mouse and keyboard
while increasing their productivity.
MICROSOFT VOICE RECOGNITION
Schools-Teachers can use speech recognitions to
improve student’s second language.
Offices- People send email and do their projects
efficiency by speech recognition.
Research Center- Scientists improve
productivity by speech recognition.
Military-Commanders can control any
equipments easily and safety by speech
recognition.
MICROSOFT-FEATURES
Commanding "Say what you see" control applications and tasks,
such as formatting and saving documents; opening and
switching between applications; and opening, copying,
and deleting files; and browse the Internet by saying
the names of links.
Correction Efficiently fix incorrectly recognized words by selecting
form alternatives for the dictated phrase or word or by
spelling the word.
Interactive
tutorial
The interactive tutorial teaches people how to use
windows speech recognition and teaches system what
your voice sounds like.
Personalization The system keep adapting both your speaking style and
accent continually improves speech recognition
accuracy.
PROBLEMS AND CHALLENGES
FOR MICROSOFT
Problems-
1. Voice distinguish
2. Command’s error on system
Challenges-
1. How to develop a new smart system
2. Strengthen distinguish system
3. Operating speed.
COMPARISON
Managing Innovation
Potentially disruptive technologies used in a sustaining
innovation framework (ecosystem)
Apple
Product ecosystem- iPad
Search ecosystem- Android
Microsoft
Windows ecosystem- Office Products
COMPARISON
KLINE: “SHARING THE CORPORATE
JEWELS”
"Strategic licensing is emerging against the backdrop of intensified efforts by
corporate America to maximize the return on its intellectual property assets,
which now account for 50% to 70% of the market value of all public
companies.“
“To judge from the results of such initiatives to date, the most powerful
benefits are economic. No company demonstrates this better than IBM, which
earned an astounding $1.7 billion from technology licensing in 2000 alone.
These revenues came with a 98% profit margin and accounted for roughly
20% of the company’s net income in that year.”
Imagine the possibilities Artificial Intelligence applications could have in this
regard.
THE FUTURE OF AI
Another DARPA creation, the internet, was in a similar
position, not too long ago.
Tim Berners-Lee’s World Wide Web
Netscape Browser
AI also needs the concurrent development of enabling
technologies, like: a semantically linked web, populated
with a web of things, and robotics.
Until then, this space is best doing a lot of the same as it
is doing now until such time as the disruptive technology
finds a model that can make it into a truly disruptive
innovation.