The thrust of Prof. Knight’s talk is that the number of atoms in an organism now is of the same order as the number of logic gates on a VLSI/CPU.

Now, in the VLSI world we have (open-source) compilers that allows us to program every logic gate, (open-source) emulators that allow us to predict the function of every logic gate, and (open-source) debuggers that allow us to inspect the state of every logic gate.

Everyone is familiar with the synergy of combined design/simulation/inspection capabilities in VLSI — so familiar that it is easy to take for granted the power of this convergence of capability.

Knight’s talk recognizes that this same convergence of capability is occurring in biology. Obviously, the inspection and simulation aspects are harder in biology — because molecules are quantum machines, the atomic-resolution sensing and dynamical simulation press against quantum limits. But these challenges are being met.

The resulting convergence of sensing, simulation, and design techniques is arriving considerably faster than (IMHO) many people foresaw. At the 50th ENC, Peter Schulz (of Scripps) gave a talk on recent developments in synthetic biology/biochemistry that was absolutely eye-popping.

A major challenge for universities is not to be left behind … it is mighty challenging to do research at this pace and at this scale, and yet still respect academic traditions of openness and commitment to teaching.

AFAICT, the key techniques for synthesizing organisms de novo are open/off-the-shelf. Good … this biological openness is necessary to the continued openness of the overall scientific enterprise.

But what’s (increasingly?) *not* open/off-the-shelf in biology are the techniques (instruments/software/algorithms) for “debugging” the organisms synthesized … these debugging tools vary greatly among enterprises, and commonly they are closely held within each enterprise.

Just to point out the obvious, effective “debugging” tools are similarly central to biological enterprises as they are to (say) software enterprises. Conversely, lack of access to effective debugging tools is an enterprise show-stopper—particularly for open enterprises.

It’s surprisingly hard to find out much about these tools. However, one shining example of openness is the GNU/LINUX community, where debugging tools are similarly open (and similarly important) to the kernel/compiler tools.

Doesn’t Drew Endy’s post (about the OpenWetWare movement) have an element of irony in it?

Namely, many MIT students trained in an open (university) environment have ended up working within the (private-sector) enterprise Ginkgo BioWorks.

In the context of an ongoing literature review of large-scale simulation mathematics, I am finding a similar trend. Namely, humanity’s most sophisticated algorithms for large-scale simulation (in finance, engineering, and biology too) are increasingly held in private hands.

Not just the software is privately held … the simulation algorithms themselves are secret.

This trend makes me uneasy, partly because it increases the power of faction, and partly because it is not clear (to me) what humanity’s shared knowledge base is evolving to become.