Plague Year by Jeff Carlson Page B

challenge.
    First was how to power something so abysmally tiny. Ruth’s teachers had called this the Tin Man Problem—if we only had a heart. Dozens of possibilities existed using synthesized fuels, proteins, live current, heat. The trick was to dedicate as little capacity as possible to energy storage and/or generation.
    Second came the Scarecrow—if we only had a brain. Nature’s oldest, most fundamental intelligence was based on chemical reactions like those of RNA and James’s amino acids, simplistic and neat, enough for some biotech, but it was a real chore to bestow the faculties of awareness and decision upon machines this size without crimping their operational speed.
    The third problem, known in polite company as the Wicked Witch, was how to create enough nanos to accomplish a goal of any worth. Manually assembling one gear composed of five hundred atoms could take a person sixty hours, depending on the material and equipment used. Automation might accelerate the process but it wasn’t economically viable, spending millions of dollars to build factories to build the nanos.
    A leading school of thought had been to bed the Scarecrow with the Witch. Nanos capable of fulfilling instructions should also be able to assemble more of themselves. Their function was their form. Once again, the infinitesimal scale had hindered efforts to master this approach, but crude kilo-atom prototypes had been doing it since before Ruth entered college.
    No single aspect of the locust was revolutionary. What made it so efficient was how well it had been put together.
    For a power source the locust used the body heat of its host, which required only a few receptors at key points in the locust’s structure. As for a brain, the locust’s creators had overcome this hurdle by dodging it altogether. The machine was remarkably straightforward. It infested warm-blooded tissue because it was unable to function in any other environment, and it assembled more identically limited yet aggressive creatures because it had been told to do so. Period. Everyone agreed that the locust as they knew it was just a test model, and yet Gary LaSalle wanted to adopt this method for his ANN.
    What a joke. Igor, fetch me a brain! Ruth must have taken her ribbing too far, though, because two months ago LaSalle had quit talking to her on the radio.
    He was right that the locust functioned quickly because it lacked complex instructions, but the man was a complete boob if he thought they could sweep the planet clean with an ANN lacking discrimination. The job was too big, the battlefield too varied. More importantly, out there in the world, below 10,000 feet, the locusts had no more hosts and would be in hibernation. They were inert, inactive targets and even a slowly replicating ANN would eventually destroy the vast majority.
    The idea was simple: release their best work, then wait and watch. But who would be the savior?
    LaSalle’s ANN, more like a chemical reaction than a machine, was composed of oxygen-heavy carbon molecules intended to bond the locusts into nonfunctional, supra-molecular clusters. Fast and dirty. James had helped pioneer the process, “snowflaking,” before declaring it unstable—and yet LaSalle’s ANN remained the smallest and the quickest to replicate, a fact he’d constantly harped upon when he was still trying to enlist Ruth’s help.
    Another faction, perhaps the most ambitious, imagined a parasite ANN that would deliver new programming to the locusts, take advantage of the locusts’ extra capacity, and turn the damned things against each other. This group was still cranking out diagnostics and computer simulations, however, and no one else believed they’d advance beyond the planning stage.
    Ruth belonged to the third team, which consisted mostly of techs with military and government backgrounds like her own. They had constructed a hunter-killer whose entire life cycle was based on disassembling locusts. A true weapon. It would