It can be hard to judge progress towards atomically precise manufacturing (APM) because the myriad research paths that contribute key technology, such as advances in atomically precise molecular self-assembly, aren't aiming specifically for APM. Therefore, scientists in these fields don't label their research "APM" or even "nanotechnology".

The task is ultimately one of design and systems engineering, not the inquiry and discovery that is the focus of most scientific research.

A chemist might discover new ways to synthesize useful organic molecules. A protein engineer might design nanoscale devices and use them to build a wider range of machinery than that found in cells. Both efforts could provide better ways to create structures with atomic precision, but they won't automatically produce a factory in a box that can make any product you can buy today, free of defects, at very low cost,* without producing hazardous wastes, and without the overhead of shipping (among many other benefits).

‍

Coordinated Effort Around a Shared Vision Works
‍

The APM situation has parallels in the artificial general intelligence (AGI) field. With AGI, the goal is to develop a system with the same cognitive abilities as those of a human, so we'd do well to understand how humans solve problems. And so we study the brain's visual system, extract principles of pattern recognition, and use that to improve systems that can search for images in large databases. We develop neural networks--mathematical models based loosely on how the brain works--to perform speech recognition.

There are numerous other examples of how AI researchers get computers to perform tasks that previously only humans could perform. But wait... image retrieval and speech recognition systems can't do the same high-level reasoning and planning that humans do, so they're not AGI. They're very useful, to be sure, but they're a bit off the path.

APM similarly consists of research areas that explore applications in disparate areas like semiconductor device fabrication and medicine, but none of these by themselves will ultimately lead to APM.

Some organizations have made it their goal to develop AGI,** and not just systems with specific capabilities like recognizing objects and understanding speech. APM has similar initiatives, like Foresight Institute, but not with same level of funding as AGI.

‍

* Something like $15 for a car, or $5 for a billion-core laptop, in 2022 dollars.

** DeepMind, OpenAI, and Numenta, for example.

‍

Further Reading

Drexler, K. Eric. "Radical Abundance".

‍