News Aggregate: Moonshots!

So in the last post I mentioned that the BioGenome Project was called a "moonshot", turns out there's more moonshots. Quite a few! Makes me wonder how many moons do we have... Am I on Jupiter??

Chemistry

Martin Burke is moonshooting for automated chemistry synthesis. Most chemicals are made very inflexibly. Basically, chemists start with a chemical to synthesize, then do a bunch of trials and errors until they get a synthesis receipe that is just cheap and with a high enough yield to be useful.

Human chemists planning a synthesis tend to use a technique called retrosynthetic analysis. They draw the finished molecule and then pick it apart, erasing bonds that would be easy to form and leaving fragments of molecule that are stable or readily available. 

Example: synthesis of Ibuprofen

Burke wants to turn chemical synthesis of "small" chemicals as easy as printing. This would make new synthesis cheap and fast.

It would be based on his 2015 machine, but on a far larger scale.

It turned out that just 1400 building blocks would suffice to synthesize 75% of all natural product "chemical space," which includes related compounds not made by any organism. "This suggests it's a bounded, solvable problem," Burke says.

Neurotechnology

Basically, the aim is to measure activities of neurons in the brain using very small sensors. The 2014 state of the art is measuring 1024 BOPS (brain operations per second) in the mouse brain, in Nanofabricated Neural Probes for Dense 3-D Recordings of Brain Activity (2014).

On the left, each needle has 64 probes arranged around the golden tips.
On the right are shown electric recordings of some neurons in a living mouse brain.

More cool pictures.

This is contrasted with the 75 MegaBOPS of a whole mouse brain, and 100 GigaBOPS in the human brain.

There has been a Moore's Law for brain measuring. Back in 1960s, scientists could measure the activity of a single neuron in the brain, and it has been very slowly rising 1000 fold per 60 years. At the current rate, it would achieve whole-brain measurement in... 2180. This is way too slow!

Michael Roukes wants to accelerate this by putting in nanoparticle "reporters" in the brain, which emit light in reaction to local neural activities, then insert needles with light receivers to receive the light signals. 

This talk was in 2016, an update was in 2017. 

But this is a very important project, for it will make Brain Machine Interface closer. Light can both be used to measure neuron activities, and to control them, using optogenetics.

Pretty much anything

Google's got a moonshot factory that is trying to make more moonshot companies. The best one so far is the self-driving company Waymo, but some others like the internet balloon Project Loon are also happening.

The only constant among its projects are:

1. it must address a huge problem; 
2. it must propose a radical solution; 
3. it must use a feasible technology

So it must be big, dramatic, and possible. Perfect.