Thoughts for a new decade

I can remember talking with Idealab’s Bill Gross just before the Millenium about themes for the coming decade.  At Idealab we decided on more internet (Bill having created Citysearch and Overture among others in the prior decade), energy and robotics.  A number of excellent companies were spawned as a results of this, including eSolar which is now putting solar energy into the Southern California grid, Evolution Robotics, about to announce some exciting products at CES, and Snap, Perfect Market and Insider Pages (since sold to IAC).

In mid-decade,  I helped Josh Kopelman as we started First Round Capital.  Josh and I had similar thematic discussions, and decided that the space between angels and VCs could provide real opportunities to invest, and that the Web 2.0 and social media phenomena should be areas of extreme interest.  I am proud to be part of First Round’s growing success, and our recent sale of Mint to Intuit, and our growing portfolio  serves as evidence that we got at least something right.

Now, we’re about to start a new decade, and I’ve been pondering what themes will make sense as we move into the teens.  As readers of this blog know, I have a tendency to invest at the way too early stage. This tendency is fortunately tempered by the various wonderful people I’ve been able to partner with, so my current thoughts should not be seen as a blueprint for First Round, Idealab, or any specific investments.

It is clear to me that one of the great revolutions in the next decade will be the exploration and exploitation of the personal genome.  Further this will largely, but not exclusively, be done using the tools of information technology that will later drive various pharma and other technologies. There should be lots of opportunity to find and invest in companies that lead and ride this wave.  First Round already has a small toe in this water.

Second, I have always found it useful to think about what could happen when a constraint is removed from present day computing.  In the decade just ending, one of those constraints was storage, which cost an enormous amount at the start of the decade, when the concept of a terabyte was large, and moved towards free by the end of the decade, when I could buy terabyte drives at Costco for $150 or less.  Wireless bandwidth will move in this direction over the teens , whether via WiMax or other technologies.  This means that all sorts of applications that need to move large amounts of data to and from servers will be economical.  As a corollary, mobile apps will become the norm, as the iPhone has shown the way.  I’ve even been watching Stanford’s CS 193P (programming the iPhone) course during this vacation. It’s available on iTunes University.

Third, I’m beginning to see people start to deal with processing massive amounts of data and making new sense out of it. The tools developed for handling giant datasets are important (First Round’s AsterData provides an example), but so are the applications. In particular, companies such as Neurovigil, making sense out of brainwave patterns, FitBit, aWare Technolgies and others making sense of accelerometer data, and other using vital signs, can lead to a revolution in personal health.

Finally, having been a participant in the eBook world as Chairman of Franklin Electronic Publishers, which sold the early Rocketbook, the eBookman,  and then sold it’s ownership in Mobipocket to Amazon, as a key part of the Kindle’s software technology,  I would say this is the decade of the eBook. And as we’ve watched the various music discovery and other sites grow during the past decade, we’re likely to see many new companies created in the eBook space – both better and more interesting hardware, and in software.

In any case, I’m going to enjoy watching the decade unfold, and will keep looking to be just a little too early, though maybe not way too early in my investing. 

Happy New Year to all.  #FRC

Comments

[Jon Beverley]

Goodness me. Ten years ago today, what was then the 'Snaffle It' team were busy celebrating not just a New Year/new millenium, but also having just received a term sheet from idealab. How time flies. Happy New Year!

[HLMorgan]

Ah yes, Snaffle It, then scan.com was a classic way too early. SMS did not take off in the US for several years, during which, unfortunately, the dotcom boom came to an end. But is was certainly right, as the number of clones now succeeding show.

[Bob Brisco]

Hi Howard, brilliant post. I love the observation that you have “have always found it useful to think about what could happen when a constraint is removed from present day computing.”

As bandwidth and mobility constraints fall away for most apps, alongside ongoing speed and storage gains, I think we experience exponential change. The new constraint becomes primarily software and analytics, such as emerging work in bioanalytics, industrial analytics (inclusive of narrow ai), consumption analytics, and analytics infrastructure. Looks like First Round is ahead of the curve again.

It should be an exciting decade!

[Michael Kuperstein]

Hi Howard. I agree with all that you said, but would add a few more mini-revolutions that I think will happen in the next decade. My perspective not only includes your view of "way too early technologies", but also the even earlier "science that is ready for prime time".

There is increasing public will and lots of investments being made in new or improving energy sources, whether it be battery technology, synthetic bio-fuels, clean coal, fuel cells, solar, nuclear, water currents or geothermal. This will spur a huge social change in what we drive, how we heat our buildings and source of jobs in new industries. I think 2010 will already be a big year for the start of electric cars in our culture.

Along with increasing wireless bandwith and a disappearing data storage expense, there will be increasing integration of all of our widgets. Distinctions among TV, PC, cell phone, tablet, displays, cameras will blur considerably with computing widgets in every room, vehicle and building we inhabit. They will become part of one giant human support network for communication, security, knowledge and control in our daily lives, whether it be a communication grid among cars for traffic and safety improvements or a sky grid among airplanes that will add many more local airports or video calls among teenagers on their HDTV displays.

There will be great medical advancements from stem cell technology and synthetic biology to cure many cancers, replace injured body parts and help us live longer.

I think the next decade will also force us to confront the shortage of clean water in our world which will spur new products from existing and new technologies in water collection, distillation, storage and processing. I see a coming revolution in food distribution. Bio-engineering of new foods will help feed the world's hungry and the creation of crop harvesting farms in buildings will localize food distribution and improve the quality of our food while reducing their transportation costs.

Every decade in the last 100 years, futurists have declared the coming of the robotic age, but having been in robotics, I don't see much progress in the last 50 years. Orders of magnitude increases in chip density, speed and storage has led to only a marginal improvement in how robots and robotic widgets affect our lives. Put simply, most robots are just computers that move as they are told maybe with video feedback and a few other sensors. I am hoping that in this coming decade, the automation industry will finally fill in the last 2 most important features that will enable a robot revolution. These are adding high density tactile sensation and adding truly adaptive learning so that robots will finally learn increasing complex tasks by trial and error by themselves. The DARPA arm is already helping military veterans that have lost their arms with dexterous manipulations like eating and using tools. The obstacle is no longer going to be hardware, but rather a required paradigm shift in producing automated behavior. Decades of brilliant minds have shown that rule-based programming which is what almost all software is, is just not good enough to make a generalized human-like robot that can help people with tasks like washing our clothes, cleaning our bathrooms, supporting the lives of the elderly and handicapped, and replacing people in most manual labor in factories from sewing clothes to general assemby of most products. A new approach that mimics how people learn will need to insert a different type of software into computers that not only learns by itself but also learns to know how to learn new tasks.

[Linda Stone]

Great post, Howard. Quantified self technologies of all kinds will play a key role in this new decade.

[twitter.com/mikehartCXO]

I agree that one of the great revolutions will come in the exploration of the personal genome with one caveat being the more it becomes politicized the less progress will come. We lost almost a decade of research on stem cells because of political agendas. Medicine continues to be one of the most highly regulated businesses where significant advancements in medicine sit idle due to regulatory oversight. The key to changing this environment is to put more of the risk decision in the hands of the patient as opposed to the government telling you what is best.

[SharelOmer]

New decade...need a lot of vision to see it, write about it, and try to read it :)

thank you for the inspiring post.

[Matt]

mobile, social, & real-time

Love that trio. I think your post hits on all these areas.

[ekolsky]

Great post,

There is an them that is present in all of these "predictions" but that you have not touched on deeply: the management of the underlying data. Yes, you can get a terabyte for $150, but what do you do with one TB of data that is useful (storing it is not useful). how do you analyze it / manipulate it / and use it near-real-time so that it helps with its intended purpose? how do you quickly validate the worth of the data -- even before you analyze it -- by looking a origin / rating / reputation and other similar areas?

I think that 1/3 to half way through the decade we will notice that data management is the key missing element in everything we are doing and seeing -- and focus on that.

then again, this may be too early even for you :)

HLM: I agree that the metadata is also of importance. And I did mean that the data should be processed in as near real time as possible, so your comments are right on.

[animation]

Great thoughts man,
I think we experience exponential change. The new constraint becomes primarily software and analytics, such as emerging work in bioanalytics, industrial analytics (inclusive of narrow ai), consumption analytics, and analytics infrastructure.

[budoushi]

In any case, I’m going to enjoy watching the decade unfold, and will keep looking to be just a little too early, though maybe not way too early in my investing.

[Keyword Software]

This is very educational. Keep on posting something that would interest the majority of people. Especially related to technology and politics.

[Monster Beats By dr. dre]

Oxycodone extraction. Difference between oxycodone and morphine. Oxycodone hcl 5 mg. 80 mg oxycodone. Oxycodone while pregnant. Oxycodone.

[Leo]

studios have better sound. its more clear. the solo hds have a ton more bass in them. the solo hds do sound good as well. the solo hds also i think are more noise caienlcng cuz if you get a good ceal/fit on your ear, they are better. the sutidos are more comfortable but there is little to none in noise cancelation. they arent really worth it. get solo hd. but get them at bestbuy and buy the $20 protection plan so if they break you can replace them. hope this helps. p.s. i have studios

[Monster Beats By dr. dre]

Every decade in the last 100 years, futurists have declared the coming of the robotic age, but having been in robotics, I don't see much progress in the last 50 years. Orders of magnitude increases in chip density, speed and storage has led to only a marginal improvement in how robots and robotic widgets affect our lives. Put simply, most robots are just computers that move as they are told maybe with video feedback and a few other sensors. I am hoping that in this coming decade, the automation industry will finally fill in the last 2 most important features that will enable a robot revolution. These are adding high density tactile sensation and adding truly adaptive learning so that robots will finally learn increasing complex tasks by trial and error by themselves. The DARPA arm is already helping military veterans that have lost their arms with dexterous manipulations like eating and using tools. The obstacle is no longer going to be hardware, but rather a required paradigm shift in producing automated behavior. Decades of brilliant minds have shown that rule-based programming which is what almost all software is, is just not good enough to make a generalized human-like robot that can help people with tasks like washing our clothes, cleaning our bathrooms, supporting the lives of the elderly and handicapped, and replacing people in most manual labor in factories from sewing clothes to general assemby of most products. A new approach that mimics how people learn will need to insert a different type of software into computers that not only learns by itself but also learns to know how to learn new tasks.

[VAS 5054A]

Well, cow heart is pretty tasty! But I never much appreciated pig brains, which are a hot commodity in Germany. Or were, back when I was but a humble lieutenant. Mayhaps their tastes have changed? Probably not. lyj

[Monster Beats By dr. dre]

To make sense of all this data we are fortunate to have built an outstanding BOT. This one is not a computer, but our Back Office Team. Led by Jeff Donnon, our CFO, this group of five people did yeoman’s work, both on and off the computer, to put together the charts, graphs and data analyses that help us provide strong transparency to our investors. When you look at almost 3,000 deals each year, and tens of funding events, there is a huge amount of data to make sense of. Yet our BOT, which I doubt we could automate into a “bot”, has done this work quarter after quarter to make sense of our investment pace, decision timing, reserve planning and so many other factors. They are helping us to find the signals in all this noisy data. While the business of judging entrepreneurs and ideas will always require some key human judgments, having our BOT adds analytics that let us focus our attention on the people part of the equation, while adding the quantitative view to our process.