On Mon, 13 Sep 1999, Terry Donaghe wrote:
> Take a look at this... I found the link at slashdot.org
I will offer some comments on this, since I saw it last night and it is going to be all over the news.
Ventner announced he was working on this much earlier this year but was postponing it until the ethicists "blessed" it. Some of you may remember that I discussed this a little at Extro3 and a little more in Extro4. Interestingly enough Glen Evans who is the director of the genome center at the U. Texas Southwest Medical Center is the one who spelled out the feasibility of this at the TIGR genome conference about a month after Extro3. Robert expands on this a little bit in Nanomedicine. This is an obvious development once you have the complete sequence of a genome to work from.
What isn't clear is whether they have actually created the bacteria (which may be hard) or whether they have simply determined what a minimal instruction set (of genes) for a self-replicating machine "should" be. All of the reports I've read indicate it is the instruction set and not actually working hardware. If that is accurate, it isn't clear how easy or difficult making the hardware will be.
Now, it is useful note, since it is fresh in my mind after a discussion with Forrest Bishop last night, that we often forget the "ecology" in which nanoassemblers (or self-replicating machines) typically operate. Since they are deriving the genome from a mycoplasma bacteria, which is an intracellular parasite, they are clearly taking advantage of an ecology where they expect the self-replicating machine to be supplied with *all* of the nuts & bolts (i.e. small-molecule building blocks) needed to be self-replicating.
Much more interesting to me, is the synthesis of a bacteria (biobot) that can do self-replication in an environment that is much poorer. The *most* interesting one is a bacteria that can function in sea water (extracting all essential materials from the sea water) using sunlight as the energy source.
Since Ventner is a very shrewd individual, it will be interesting to watch how this plays out. However, bear in mind that *ALL* they have done is take a Pentium III processor, thrown out all the instructions that were added over the last 10 years, leaving the original basic computer, something like a 386. The only hard part about it was they did it with only about 2/3s of a 386 processor handbook that explains what the instructions do. I'll be fascinated to know whether they did it the hard way (figuring out what the other 1/3 of the instructions do or) or the easy way (delete them one by one and if the program no longer runs, you must conclude, "yep, *that* one is essential").
It is worth noting, that the timing of the announcement will probably make this the hot topic at the TIGR conference next week [funny how that works isn't it...]. Since I'll be there, if any of the details come to light, I'll comment on them for us.