The Next Big Thing

In Secular Decline I looked at where we are in the business cycle. What I said basically was that semiconductors and the microprocessor advance were no longer providing excess profits. To a pretty fair extent the world has been computerized. Profit opportunities are declining.

So then comes the question: What will be the next big thing? The The Online Investing AI Blog linked has a nice graph of the business cycle. But let me cut to the chase.

Solar power. Fusion power. Mind-Machine interfaces. Nano-technology. Personalized medicine. Rapid prototyping.

They also mention "Smart AI-powered investing for the masses". Now I don't see how that can work. If they can identify above average opportunities reliably, the masses will all invest and there goes your above average opportunity.

As to the others. There are lots of opportunities in nano-technology. Which one? And yes, solar is probably one or ten breakthroughs away from being low cost enough to start capturing a lot of the electrical energy market. Except wind is currently lower cost and the cost reduction curve is more reliable. I can tell you, almost certain, that wind will become lower cost than coal fired electrical plants when wind turbines reach the 8 to 12 MW (peak) size. Right now the 5 MW (peak) size is just going into series production. That means that at the best wind sites the cost is below coal and at the worst it is above coal.

There are two problems with wind. All the best sites are in places (like North Dakota) where grid connections are sparse. And wind is intermittent. Which means that without storage its contribution is limited to about 10% of grid power.

Which points to two investment opportunities. The first company (GE? Westinghouse? Siemens?) to come out with 2 MV DC transmission equipment and 2 MV DC to AC conversion equipment (AC to DC is easy) will help bring wind from the upper Mid-West to the loads in the more populous states. It also gives rise to the possibility of wind going above 10% to perhaps 20% of grid power because of both wider generation averaging and load averaging.

So what is the second opportunity? Low cost very high power energy storage. Will it be batteries? Fuel cells? Flywheels? No one knows. What we know is that the technology will have low turn around losses (generation - storage - generation), very high energy capacity, reasonably long life (5 to 20 years), low losses over 24 to 72 hours of storage, and low cost per KWh stored (below 2¢ per KWh with the possibility of getting below .5¢ per KWh). A pretty tall order.

As for solar - because it is better matched to grid demands (high in the day low at night), and because of its greater predictability (when the sun shines) it can probably go to 20% to 30% of grid power before storage is a necessity.

In addition wind peaks during the low demand season of winter and sun peaks during the high demand season of summer. So the two forms of production are to a certain extent complimentary. However, solar peaks at noon and the load (air conditioning mostly) peaks at 3 PM. So economical 4 to 6 hour storage would be good. However, that storage would be mostly idle in the winter except for load leveling.

So what else will be needed? Smart grid equipment which can turn on and off loads like electric water heaters, refrigerators, and air conditioners to match supply and demand better over short intervals (say 15 to 30 minutes). If plug in hybrids become big they could be used for load leveling over short intervals as well allowing motor fuels (cellulostic ethanol?) to be used to arbitrage the high cost of day time electricity with the low cost of night time electricity. Everybody will become demand metered. And that is another opportunity.

Fusion power? There is no working prototype yet. Personalized medicine? Another 100 to 500 breakthroughs are needed. One of which is much lower cost genetic sequencing and the ability to produce required molecules in mass quantities (1 to 5 grams - which is a lot if you are making them a molecule at a time). Rapid prototyping? I saw a machine for sale a month or two ago that fits on a desktop that costs $5,000. A little pricey for a home that might only need 5 to 50 custom parts a month. And only good for small plastic parts in any case. To make such a system work you also need to be able to make metal parts too. Which means a small milling machine and a small lathe. Each of those is going to run another $5,000 or so with current technology and auxiliary equipment like clamps measuring tools etc.

Mind machine interfaces? Another 1,000 to 5,000 breakthroughs required.

Now what can we say about all these things? Progress is being made and some day in the next 5 to 20 years each will cross the 2% penetration threshold where they take off rapidly. OTOH there is danger too. If you can make your own molecules in a well stocked home lab, virus making is just a few computer codes away.

How will man, with his hunter gatherer brain, with all his passions, survive such access to power? That is the $640 trillion dollar question.