Below are some people who I've supported with small amounts of funding. I've given away more than $100,000 so far. This money, historically, has come from Astera Institute, Experiment Foundation, and ResearchHub. I typically advertise new funding opportunities on my social media pages.

Hyperspectral Biology ($60,000)

The human eye is both remarkable and pathetic.

Remarkable because the lens reshapes itself in a fraction of a second, letting us focus on objects near and far. The fovea, a pocket of cells at the back of the eye, also has three types of cones, which can together distinguish millions of shades of color.

And yet, those three cones can only distinguish light in a narrow band of wavelengths, stretching from about 380 to 750 nanometers. Pah!

A hawk has four cones! It can see ultraviolet light! Its cones are also far more densely-packed — roughly a million per square millimeter — so a hawk can spot a mouse in a field from three times the distance a human could. (Not that you'd care to spot a mouse from a hot air balloon.) Mantis shrimp also have 16 photoreceptor types, 12 for color, enabling them to see wavelengths of light from 300 to 720 nm.

Clearly, biology is more complex — and beautiful — than human eyes can appreciate. Despite our inadequacies, though, we still describe the natural world with our own biases. If the human eye could instead perceive the full spectrum of light, we'd not only see more of biology but could also build amazing things.

Every molecule interacts with light in a unique way, absorbing some wavelengths while reflecting others. Every molecule, then, has a unique “spectral fingerprint.” If our eyes could see a wider spectrum of light, we could — for example — look through a microscope and immediately tell a cancer cell apart from a healthy cell. We could even see methane gas floating in the air, as the molecules have carbon-hydrogen bonds that absorb infrared. For generations, though, we had no way to see most of these signals.

But in the 1980s, NASA engineers invented the hyperspectral camera. Unlike a normal camera, which collects only three bands of light, hyperspectral cameras collect a full spectrum for each pixel in an image. Each pixel, in other words, has a full spectrum attached to it, quantifying how much light at each wavelength it absorbed or reflected.

Hyperspectral cameras do all kinds of things. In the food industry, they search for bacterial contaminants. As food rolls off the processing line, a hyperspectral camera looks for microbial signatures. In one study, a hyperspectral camera quantified how many E. coli cells were in packaged spinach nearly as accurately as plating and counting those cells on a petri dish, with an R² of 0.97.

Hyperspectral cameras are also used for in ovo sexing, or to figure out the sex of eggs before they hatch. Roughly seven billion one-day-old male chicks are killed each year in the egg industry, simply because they don't lay eggs. Hatching factories place baby chicks on a conveyor belt shortly after they hatch, sex them, and discard the males into a metal blender.

A German company, Agri Advanced Technologies, built a machine called Cheggy that uses hyperspectral cameras to sex eggs before hatching. The machine lights up eggs from below, using a halogen lamp, and then uses a hyperspectral camera — positioned over top — to look for feather signatures within the egg. Female chicks in brown eggs have a unique pigment, visible through the shell, that distinguishes them from males. The cameras are about 95 percent accurate and can sex 25,000 eggs per hour, at day 13 of incubation — right about when researchers think an embryo may first process pain. The other option is to puncture the shell, draw allantoic fluid, and run PCR for sex chromosomes, but that takes much longer and is more expensive.

Hyperspectral cameras are even mounted on satellites, used to study algal blooms, quantify soil moisture for crops, and predict the locations of mineral deposits. (Micas and iron oxides absorb infrared light in unique ways, which can help mining companies decide where to drill.)

Still, we've barely scratched the surface of “hyperspectral biology,” or how we might use hyperspectral cameras to study biology.

To grow this field, I recently awarded nearly $60,000 in microgrants to nine people. Some recipients are collecting datasets on how molecules interact with light, which might prove helpful for training machine learning models to differentiate those molecules from background signals (like sand or soil) from satellites. Other recipients are simply pointing hyperspectral cameras at things nobody has yet measured. For example, a student in the UK will tour museum collections to photograph bird plumage with hyperspectral cameras, studying whether the feathers of a species look measurably different before and after the Industrial Revolution. Each recipient has committed to sharing their data and results publicly.

These microgrants are not large enough to do a big experiment. The funds are basically enough to buy a camera, say, or to dabble with an idea. My main hope, then, is that these funds will “nudge” people to work on hyperspectral biology who wouldn't otherwise. Their excitement may start with $5,000, but if even one person digs deeper and wins a larger grant for their work down the road, then I'd consider that a major success. All the grant winners are listed below.

Thanks to the Experiment Foundation for making this possible.

Fast Biology Bounties ($13,500)

Wet-lab biology is a major bottleneck for scientific progress. Atoms are harder and more expensive to manipulate than bits, which means a task as routine as cloning a single gene still takes days, hundreds of dollars in reagents, and tens of thousands of dollars worth of equipment. With support from Astera Institute, I offered $10,000 in prizes for ideas to speed up or cut costs for common wet-lab methods, judged purely on originality and technical tractability. Below are some of the winners.

OpenFlower ($10,000)

A microgrant to support Nick Desnoyer and his work on OpenFlower, an effort to use genetic design tools to create new types of flowers.