LOL @ "Accepting there's a chance" - There's documented evidence. It has happened. It is happening. It will happen again. There are ways and means of controlling access, but that didn't seem to do cisco, juniper, dell, huawei, WD, maxtor, samsung, gemalto etc much good from falling prey - companies like microsoft, intel, ATi just bend over and ask it it can come any harder. And it's not just the big the firms, the little ones get hit too.
And no, not the future, this is a project for now. That's how you make thing happen.
The technologies required exist now, we just have to package them together, and write some software. And, yes, that development starts on Earth. It can sensibly be distributed too. It's a little bit of work, but it's not beyond the realms of possible.
Starting with a CNC milling tool(larger concern is debris in microgravity, not impossible to solve), a few 3D printer, preferably developing a method that can use one system for multiple materials, but if required just several types(some varients are more suited than others), a couple of delta robots, a few mechanical arms etc, -=- Preferably you would be considreing generation, but early iterations can be grid based -=- and you start building small scale industrial capacity that is capable of building the parts required to build itself. You then make it print off higher quality cogs, gears, etc to be able to upgrade itself to higher precision, then you start expanding the physical capabilities of the seed factory. As that starts, you could also begin an intermittent run on a clone then ship them parts to a lesser fortunate Asgardian with assembly instructions. I'm sure to not be the only one to attempt such a feat, and once I've open sourced plans there will be even more. If such a thing would be of a concern to you, then also at about this stage it should be possible to make it pay for itself by way of renting it's capacity to the public and or private companies. Basically, from feeding it materials a wide range of goods should be possible to be produced - we can additionaly start an archive of open source plans - that should, apart from a few things required for purchase, be able to further enhance itself with more heavy industrial processes. Once there's a sensible number of units with high enough quality output through series of self improvements a kit can be produced distributed and shipped to near launch site for assembly. Launching it at this phase in it's cycle should keep it's weight down, and still allow it to upgrade itself autonomously in orbit.
And now for why it's relevent to this topic.
One of the things it would be sensible to upgrade itself with whilst it's up there is chip fabrication capacities - then it'd make sense having a reflow oven and thinking about PCB fabrication. My ground unit will have PCB fabrication, pick n place, reflow oven already automated well before it's getting lifted and one of the next obvious steps to devel would be chip fabrication capacity, I could do with some of that. Most of it would be built at material cost, too. It'd make sense if possible to utilise methods and procedures that would then directly transfer to microgravity application -=- Then we can upload the plans to the orbital seed and build that mostly out of LEO debris. If we can devel this before we even attempt to lift it, we could actually print off the chips, boards, and shielding ourselves and maybe stand a chance of retaining control over it whilst it's up there. By distributing the production we can produce rapidly. It will also distribute cost, and could in theory have the entire enterprise dropped down to materials cost (which could be paid for by the machines) and launch costs.