Biotech is "borrowing" a few tricks from embedded development. Trying to get an LED to light up to prove your code is running - similar to what they've done with triggering fluorescing genes. This is sort of like writing sentinel values when interrupt routines get called and reading them out after the fact.

Once you have these tools in place, building the fundamental blocks to bring a system up is much easier.

Yes, when I first saw the fluorescing genes, I couldn't help but think "Ah, they're up to printf debugging. Awesome." (No sarcasm. It is awesome.)

Genetics strikes me as a lot like computer vision or AI. A lot of really wild promises about 20-30 years too early, then when the real stuff starts to happen everyone is long since tired of all the hype.... but that's when the real stuff starts to happen.

I work in bioinformatics and have a reasonable amount of experience in the lab. In my perspective, scientific fields like computer science and biology progress in opposite directions. We are starting to see the intersection of the progression.

In computer science, a strong artificial foundation is built and we abstract from that foundation to get complex and useful results. In life sciences, we have the complex results and we are trying to reduce it to get at the foundational elements.

Computer science fields like computer vision or neuro-networks are progressing towards complexity that biologists might regularly see in their research. And genome editing technologies are progressing towards the fundamental building blocks computer scientists might normally expect in a system.

Amara's law: we overestimate the impact in the short run, and underestimate it in the long run.

> A lot of really wild promises about 20-30 years too early, then when the real stuff starts to happen everyone is long since tired of all the hype.... but that's when the real stuff starts to happen.

Here's a paper showing how to make fluorescent reporters from 1995: http://www.ncbi.nlm.nih.gov/pubmed/8777060

But I agree, the real stuff is a long-term accumulation of studies building on each other. The reporters were (and still are) interesting/useful for people in molecular biology... but their application is only now becoming interesting/useful for others.

So like console.log for cellular biology?

Except the only thing you can log is '.'

Is there a reasonable way to start playing with crispr? Can I order stuff to do it with?

Sure, you can do it. Many things are available on Addgene or a friendly researcher might share genetic material. You still need lab equipment, so you need either a public science lab or quite a bit of money.