Bob Metcalfe, the internet pioneer who has formulated what is known widely as Metcalfe’s law, and who now calls himself a venture capitalist, has turned his attention to energy. In a presentation which he says is his first-time use of PowerPoint, Metcalfe outlines why and how we should learn from internet history to solve the problems of energy.

Bob Metcalfe: Internet History Applied To Solving Energy

“We used the Internet to build the Internet, so let’s use the Internet to solve energy.”

Internet history shows the best innovation vehicles are competing teams of research professors, graduating students, scaling entrepreneurs, and venture capitalists.  The Internet-based Enernet will be intelligent, layered, distributed, standardized, networked, asynchronous, symmetrical, and it will have storage.

A side-by-side comparison of enernet and internet

Voice – Video – Data

The voice, video, and data monopolies were surprised when voice became more than telephone, surprising AT&T and their FCC, video became more than television, surprising the TV networks and their FCC, and both became data on the Internet, surprising IBM and their Justice Department.

Feed – Food – Fuel

DC chose to subsidize the wrong feedstock (corn) and the wrong fuel (ethanol).  Cellulosic whatever is next. Algae are many more times productive than agricultural crops, harvested every day, not just once a year.  So,  for example, algae produce oil 400 times faster than corn, ~48,000 liters per hectare per year. And alage don’t take agricultural land (think desert) or drinking water (think ocean or sewage).

We can use the Internet for COLLECTIVE INTELLIGENCE, better to solve energy.

Do we want to turn the world’s lights off or on?  The USA does not waste energy because we are rich; no, we are free and rich because we have plentiful energy.  China, India, and the rest of the developing word want to be free and rich.  We should welcome that. Technologies providing cheap and clean energy in squanderable abundance — at last “too cheap to meter” — are half the battle.

Let’s light up the world with squanderably abundant, cheap and clean energy.

The Internet Movement’s color seems to have been TRANSPARENT, but the Ethernet and Linux sub-movements, for example, had yellow as their color (and Linux has a Penguin). Movements often have colors, and the energy movement seems to have chosen a color … GREEN.

In breaking from the past to finally solve energy, I say we’d better change colors.

Photosynthesis gets most of its energy from red and blue light.  So, shall our energy movement choose red or blue?

Blue is the new green.  I hereby move that Enernet adopt blue.

Why blue?

Of course, looking back up, the sky is blue. The vast majority of Earth’s energy (including fossil fuels) comes from the sky. 10,000 times the power we use comes from the sky. Enough in an hour to serve us humankind for a year.

Earth is blue because it is mostly oceans…

Earth’s oceans are likely to be abundant sources of tidal (gravitational), wave, thermal, wind, solar, geothermal, … and/or mineral energy.  And they have a lot to do with climate. Let’s explore the deep BLUE sea.

Energy solutions will likely emerge from the sky and oceans.

So when each Arpanet site reported destination counts of our growing Arpanet packet traffic, as counted by Imps, we left off the largest numbers, namely traffic that never left the building, which we dismissed as “incestuous traffic.” Surprise: Ethernet LAN traffic is today dominant. There will be Enernet surprises too.  I wonder how much energy will be generated and consumed all within the same house, in a sort of energy LAN?

Circa 1980, I came up with this 35 mm slide to encourage my 3Com customers to buy bigger Ethernet networks, to enjoy network effects by growing their networks past what I called critical mass.  In 1993, George Gilder, writing in FORBES, called this slide “Metcalfe’s Law.” I have been defending it ever since.

What if Enernet is really mostly a networking problem, a problem whose solutions will best be found by looking for them through a networking lens?  And if so, might Metcalfe’s Law apply to building the Enernet.  Well, the so-called “power grid” is certainly a network.

There might be a “law” about the cost of solar power ($ per peak Watt) versus cumulative production (millions of peak Watts), a “learning curve” for solar cell manufacturing costs.

There may be comparable cost decline laws for wind, geothermal, fission, fusion … let’s find them.

If Global Warming were solved tomorrow, we would still need cheap and clean energy. And vice versa.

…energy and the environment are NOT the same thing.

We hear that energy must develop public awareness, political will, and go to Washington for some sort of Manhattan Project, or perhaps even an Apollo Program. Trouble is, energy is much bigger and will take much longer than that.  More like a Marshall Plan for rebuilding Europe. Plus, the status quo runs Washington…

The last time we went to Washington for energy, we got DOE, a huge series of earmarks and government jobs programs that leave too little oxygen for energy innovation.

From Internet history I learned that the best place to put our research dollars is at competing research universities vying for government grants.

Focusing on the USA, here are 2006 energy source and usage estimates from the excellent National Academies report, “What You need To Know About Energy.”

According to DOE, more solar energy falls each day on the USA than we use in a year.  Incident solar energy is 37 * 10^16 kWh/year or 46,700 Quads/year, which is, get this ~500 times (=496.28) more than the 97.1 Quads above for 2006.  Over the whole Earth, the multiple is ~10,000.  Solar has a low floor (.07 shown), but a high ceiling (10,000x).

Look at this huge, old, and very slowly evolving system, with many pieces, levels, connections, interfaces, standards… We Enernet innovators could easily be overwhelmed, not just by the down and dirty status quo, but by the complexity of it all.  What to do?  How to manage all this complexity while evolving it?  What worked for the Internet?

Internet history demonstrates the enormous power of of LAYERING.  The Enernet should have layered ARCHITECTURE.

If the Internet is any guide, in the future we’ll have more smaller DISTRIBUTED power pants.

This may be Internet history’s killer lesson for energy: Go distributed!Not centrally generate power and then distribute it, but generate power in a distributed system.

Distributed energy (in the Internet sense) means the grid become more peer-to-peer, more multi-vendor, with more standards, more competition (FOCACA).  The transmission of energy then become more networked, and more symmetrical, more among than between.  I want my home and car energy systems to be able to buy energy from power grids, but also to sell.  Radios and TVs are being replaced by PCs, which upload, not just download.  Ditto nodes in energy grids.

Today’s electricity grid is synchronous, 60Hz-110V AC in the US, because of lighting, motors, and transformers in the days of Edison and Westinghouse.  But AC makes the grid fragile and inefficient.  By analogy, for a new Enernet, maybe we should consider standardizing an asynchronously switched grid, say switching packets of DC power.

Overplaying Internet lessons so far, I see an Enernet in 62 years

• that is mostly a distributed, layered, symmetric, asynchronous, switched DC power grid,
• with networked intelligence extending to trillions of leaves of the smart grid,
• with energy harvested and stored, off, on, and in the grid,
• from Sun’s fusion reactor, using distributed solar harvesters,
• from Earth’s fission reactor, using distributed geothermal harvesters,
• from man-made distributed and perhaps mobile fission and fusion reactors,
• networked with electricity and information, including robotransport on demand,
• with squanderable abundant, cheap and clean energy providing
• clean water, space travel, …, freedom, and prosperity for all.

These extracts from Bob Metcalfe’s presentation give the general idea, but certainly they don’t do justice to the work that went into it. So if you have the time, please do go see the original at

Bob Metcalfe: Internet History Applied To Solving Energy