Jan 20, 2020
Dr. Louis Rosenberg founded Unanimous AI to pursue a crazy idea
– that we can amplify the intelligence of human groups by
developing a new form of AI based on the biological principles
Intelligence. The idea panned out, resulting
AI technology, a powerful method for amplifying group
intelligence. It has been validated by over 20 academic papers, has
been funded by NSF, and was named “AI Innovation of the Year” at
South By Southwest.
Dr. Rosenberg began his technological career at Stanford
University where he earned his Bachelor’s, Master’s, and PhD
degrees. His doctoral work focused on robotics, virtual reality,
and human-computer systems. While working as a graduate researcher
at the U.S. Air Force Research Laboratory (Armstrong Labs)
in the early 1990’s, Rosenberg created the ‘Virtual Fixtures’ platform, the first
immersive Augmented Reality system ever built.
Rosenberg then founded Immersion Corporation to pursue virtual
reality technologies. As CEO of Immersion, he brought the company
public (NASDAQ: IMMR) in 1999. Rosenberg also founded Microscribe,
maker of the world’s first desktop 3D digitizer – the Microscribe
3D – which has been used in the production of countless feature
films, including Shrek, Ice Age, and A Bugs Life (acquired in
2009). Rosenberg also founded Outland Research, a developer of
advanced interfaces for mobile devices and augmented reality
(acquired in 2012). He has
also worked as a tenured Professor at California Polytechnic
(Cal Poly), teaching
engineering, design, and entrepreneurship.
A prolific inventor and researcher, Louis Rosenberg has been
awarded more than 300 patents worldwide for his
technological efforts in virtual reality, augmented reality,
artificial intelligence, and human computer-interaction. He
currently lives in sunny California with his wife and kids, as a
longtime vegan and friend of all animals, big and small.
In today’s conversation with us, Dr. Rosenberg explains the
dangers of traditional AI, how swarm technology works in nature,
and how we can apply it to making better decisions as a people
- Traditional AI is pushing to replace people with algorithms.
But there is a way to keep people in the loop.
- The reason why birds flock, fish school and bees swarm is
because they can make much better decisions in groups than they can
individually. It helps their ability to survive. They are much
smarter together than alone.
- Swarm AI is a new form of AI that connects people and
algorithms to get significantly smarter, make much better decisions
together than they could on their own, while keeping people as part
of the process.
- When you want to use AI to solve a real problem you have to
start with a good data set that is up to date and accurately
reflects the problem you’re looking to solve. For most problems in
the political, financial or business world it’s very difficult to
get a data set that works.
- Most of the data used is out of date almost immediately, and
it’s really only a simplification of what people think and feel.
For example, how people click and connect on the internet does not
reflect their true feelings.
- AI that makes predictions about people, what political messages
are going to work or what products people are going to want, is all
based on data that doesn’t reflect people’s true thoughts and
feelings. It’s also infused with a lot of biases.
- For example, Google and Amazon have reduced you down to a set
of data that it thinks you want, so when it wants to give you an
advertisement it makes use of this estimation. Then it becomes a
self-fulfilling prophecy where you’ll probably buy something within
its selections, which then reinforces its model.
- When this is applied to news, it becomes a self-fulfilling
prophecy and we end up polarizing a population.
- Polls are polarizing, finding and reinforcing differences.
Nature does the opposite. In a school of fish, there are no leaders
and no followers. Everyone is equal, and they make decisions about
their survival together as a system.
- Over millions of years all the schools of fish that split in
two, that couldn’t agree on which way to go, died out. The rest
enabled through this process of swarm intelligence to make
decisions that are optimized for the group as a whole.
- Whereas we humans make decisions in large groups but we’ve made
this method of polling which is polarizing, showing our differences
and reinforcing those differences. We then we become that school of
fish that splits in two or just can’t make a decision.
- In a study with Stanford Medical School and a group of
radiologists, when they worked together as a swarm, we reduced
their diagnostic errors by 33%, vs if they made these decisions on
their own or if they took a vote.
- In a study with Oxford University when a group of financial
forecasters came together as a swarm, they were able to make 36%
more accurate forecasts about the price of gold, oil and the stock
market than when they were working alone.
- A swarm creates a much better mix of the group’s insights, plus
it is interactive so individuals experience the group converging on
the answer. There’s more buy-in and they are collectively more
satisfied with the decision. Even if it’s a large group, everyone
feels like they are consequential – like they had an impact –
because they were working together and they experienced it.
- With a polling system, they put their answer in a black box and
if it doesn’t come out the way they wanted, they feel like nobody
listened to them.
- With swarming, when you have a group of people, you can make
them smarter but you can also make them wiser where they make
decisions that are really better for the whole group.
- Hierarchy hurts decision making. However, in a swarm everyone
is equal and anonymous, and therefore honest, because they’re not
trying to say what they think their superior wants to hear.
- With a swarm, the group can prioritize a set of issues in a way
that will be more agreeable to the whole group than if they just
argue about it, take a vote or poll, or try to do it
- xPrize has used the software to get their visioneers together
and prioritize the most significant problems they want to
- Diverse groups make much better decisions than monolithic
- In a study of sports, a group of diverse fans were
significantly more accurate in predicting the games than the
experts who were getting paid to do it as sports writers.
- People are smart and we shouldn’t replace them with
- Yes AI is powerful, and we can’t lose people in the process,
with their intuition, emotion and insight.
New Hope For Humans In An A.I. World
Connect With Dr. Louis Rosenberg