Sounds Too Good To Be True

If it sounds too good to be true, it probably is.

We've all heard that, and it is basically wise advice. I think it applies to a new report from Nature that says "Desktop fusion is back on the table" -- but does it apply equally well to molecular manufacturing?

Let's take a look at fusion first:

Can the popping of tiny bubbles trigger nuclear fusion, a potential source of almost unlimited energy? This controversial idea is back on the table, because its main proponent has new results that, he claims, will silence critics...

The idea is simple enough. Blast a liquid with waves of ultrasound and tiny bubbles of gas are created, which release a burst of heat and light when they implode. The core of the bubble reaches 15,000 °C, hot enough to wrench molecules apart. Physicists have even suggested that the intense conditions of this sonoluminescence could fuse atomic nuclei together, in the same process that keeps our Sun running...

Desktop fusion would be a wonderful advance, obviously. But I'll believe it when I see it.

Why, you might ask, don't I as easily dismiss the promise of advanced nanotechnology?

The reason is that I've looked closely at both the benefits and the dangers.

Here's how my thinking works:

Many new technologies (e.g. nuclear fusion) offer windfall benefits but few high-level hazards. Some emerging technologies come with very serious risks. In the case of molecular manufacturing (MM), these include economic and social disruption, an unstable arms race, and the potential concentration of overwhelming power in the hands of a few.

Other emerging technologies, such as robotics, genetic engineering, and artificial intelligence (AI), also may bring serious risk. Then why am I not as concerned about those technologies as about MM?

The answer is timing. With robotics and gen eng, the effects are likely to appear more gradually than with MM, that is, spread over a longer span of time. Thus, society will have many years, perhaps decades, to adjust. By contrast, MM could appear quite suddenly.

It is quite possible that strong AI will be just as revolutionary and dangerous as MM, and that its impacts will be equally sudden, but I think it is less clear that AI is as imminent as MM. Others disagree, of course.

Because MM will challenge us with huge risks, and because MM appears to be the earliest revolutionary new technology that will affect us, I have chosen to devote my energy to understanding and raising awareness of these issues.

Mike Treder

Tags: nanotechnology nanotech nano science technology ethics weblog blog

Comments

[michael vassar]

It sounds like you are saying that (near term?) MM may not actually be that probable, but that it's rational for you to prepare for it anyway on expected utility considerations. Is that your actual position? I thought you were fairly confident of MM developing (soon?).

Anyway, there is little reason to expect MM to be as sudden as strong AI, as the latter can self-improve recursively without the need for a human bottleneck while the former cannot. MM can bring an industrial revolution compressed to a few years. For strong AI expect to see one compressed to a few days or less. I agree however that MM looks likely to be the earliest revolutionary technology to hit us.

[Mike Treder, CRN]

Actually, I am confident that all of the technologies I named will be developed and will have a significant impact on society. Some will be here sooner than others, some will present grave risks, and some will
arrive quite suddenly.

My judgement is that MM fits all three of the above criteria and therefore requires serious investigation. Whether MM is developed in 10 years, in 15 years, or sooner, or later, does not change the fact that we should prepare now. Moreover, because the stakes are so high, it behooves CRN to use the earliest plausible development time as a target date for effective readiness.

[Phillip Huggan]

At some point the concept of a replicator will be a reality. This will probably emerge in a nanotech context regardless of whether it is specifically Drexlerian.

I don't think you have to be worried about someone using AI or Genetic Engineering or Robotics to take over the world in the same way as MM might facilitate. Sure you can wipe out the human race with a pandemic, or AGI our demise, or...engineer faulty assembly line robots and crash test robotic facilities and cause increased vehicle fatality rates (how did robotics make the short list?!).

With MM you just might be able to take over the world. At least you would be worried other might try. Fusion is just an energy source that seems to be sapping away a lot of money that could otherwise make polymer solar cells or giant bucky-rock wind turbines a reality.

[pierre basmaji]

How the nanotechnology can help the fusion Nuclear?
How we can see the difference between paific fusion and weapons enrichment? can nanotechnology materials helps in this filed.
Regards

[Chris Phoenix, CRN]

Nanotech can't directly control the fusion reaction--you can't mechanically smash atoms together hard enough to make them fuse, for example. And I calculated once that it would be basically impossible even to make a precise enough particle accelerator to help with "hot" fusion; after just a few nanometers of travel, Heisenberg uncertainty makes the atoms miss each other most of the time.

Nanotech machines could help with isotope separation. Nanotech materials might also help with reactor construction. So it could be an enabling technology. Also, if and when molecular manufacturing makes large-scale construction really cheap and fast, so we can build ITER in a month for a few M$, research may speed up.

I don't know what to say about peaceful vs. weapons uses. Any general-purpose technology can and will be used for both.

Chris