Greg is a venture capitalist and former CTO of AT&T. He graduated from MIT in '77 with a degree in Physics and obtained his PhD in Physics from Harvard. He gave some good advice on evaluating new technologies. He believes strongly in
learning curves and thinks many of the hot new technologies today could have been predicted easily by just looking at learning curves.
Here is a description of the talk:
How not to work on the wrong thing: lessons from science and businessEven smart people can choose a thesis in a field about to go out of style, or start digging canals a week before train tracks are laid. How can we think more systematically about the future, and reduce (but never avoid) dead-end decisions? I'll make a few suggestions, and try not to muddy the waters any further.
Greg Blonder's BioNotes: A lot of people are sleeping through life
The two big fields in '75 were Chemical Engineering and Nuclear Engineering but in hindsight they would have been bad choices due to events such as 3 mile island, etc.
But these events were all predictable (he didn't really explain how though)
Gave a funny story: two guys are walking in the woods and see a ravenous bear approaching at full speed. One guy starts putting on his running shoes. The other guys says, "Why are you wasting time putting on your running shoes? You aren't going to outrun the bear." His friend said, "I don't have to outrun the bear, I just have to outrun you." Moral of the story: You just have to do better than the other guy.
"The best way to predict the future is to invent it." –Alan Kaye
But he prefers this advice: "You can see a lot by just looking." –Yogi Berra
The internet phenomena had been growing exponentially since 1969, it didn't just start in the 90s.
The worst thing for an entrepreneur is to be too early or too late
VocalTech was too early; Skype was just right
Video over the Internet has a lot of potential right now
Cable providers put their own Trojan horse in your homes – the cable modem.
Takes 10-20 years for a new idea/breakthrough to become profitable
AT&T didn't pay attention to the long distance learning curves
He told Hotel chains to get rid of room phones and by a wireless repeater because everyone was going to have a wireless phone. Create a value add service such as hitting *2 from your wireless phone in the hotel and reach the front desk. They didn't listen.
LED lighting is going to replace all lighting
You should be thinking what is going to happen when LEDs are going to take over? Who will be hurt?
Electricity and gasoline follow learning curves (i.e., it is easier to make both)
However, oil does not follow a learning curve and this is disturbing. It might mean that it is not getting any easier to make it or we are burning more than we find.
Seven most important factors to determine if a new technology will diffuse through society (compared to the old technology): relative advantage, compatibility, complexity, observability, trialability, social acceptability, and regulatory.
CD vs vinyl, HDTV vs NTSC
What should you do?
* Take time to look around and wonder
* Question authority, but listen to experience. Experienced people are usually right, you have to find the times when they aren't.
* Learn "Scenario Planning" and other formal techniques for modeling new technologies
* Don't look back. You are going to make mistakes, you just have to run faster than the other guy
He never looks at a business case – he wants to hear the new insight, how they will make money, and team background
Modeling isn't a catch all – he has been surprised several times
He would have absolutely put money in Ebay, but not Google. Too easy for someone to switch to a new search engine.
7 Comments:
Once again, a good post.
I am glad that Greg mentioned about the possibility of LEDs replacing all lighting. That's the business of my $50k team. Nice...
The CEO of Color Kinetics said something similar to my New Enterprises class last term:
http://rallenhome.com/blog/mit-sdm/2005/03/george-mueller-ceo-color-kinetics.html
The LED thing was discussed in Disruptive Tech too (with mostly the same conclusion).
And the bear story is funny indeed, if not original.
But I don't buy much of the rest of his talk: most people who claim you can predict the future are, well, there are a number of words for that ;)
Sure, timing is everything, we know that already. Sure, experience helps, but there have been studies recently showing that for many items, turnover in designs and implementations is so fast that there is no time for an experience curve to bear fruit. At least he's honest about Google ;)
BTW, is the front page recent posts section updated automatically? Or does a Robbie Fan like myself need to individually check each sub-blog for entries not listed on the front page?
So what do you think is the best way to commercialize a technology?
First to market?
Time to market?
First-mover advantage has been shown time and again to not be critical in the long-term. We talked a lot about this in Disruptive Tech -- weren't you in that class??
But I think the question itself is a fallacy: there is no "best way" to commercialize a technology. You have to prepare for both the short-term (where being a first mover helps), and the long-term (with IP thickets, complementors if you're building a platform, or platform alliances if you're a complementor) especially by looking for the dominant design features and relevant process optimizations for them. No one said it's easy ;)
Thank you. That's what I have been telling my team in i-teams. There's no panacea in bringing a new technology to market. Everyone assumes that our technology is the best out there and being the first in the market in key.
No... I didn't take Disruptive Tech last semester. But I am planning to take next semester. :-)
Yoav: Each of my blogs is independent. So no, the recent posts on my home page are only for that blog. It would be nice if Blogger could handle multiple like that!
Post a Comment
<< Home