If a transforming robot is cooler than a regular robot, then what’s cooler than a transforming robot? Why, a bunch of transforming robots that combine together to form one giant robot! So how can we make that happen, and if so, what would happen?
This might seem like the hardest part, but depending on what you’re doing this is actually the easy stuff. It’s a big long engineering problem and like all engineering solutions is going to be approached as a study of “best compromises” depending on what performance targets you need to hit. We can get some sort of transforming robot ready right now, it’s just a matter of figuring out how they can combine to form an even bigger robot.
The software side can be easy or tricky depending on how complicated you want it to be (just a glorified R/C toy with someone behind a control pad or full-on artificial intelligence). Since full A.I. robots are more fun than “dumb” ones and since the Transformers themselves are fully independent A.I.s let’s assume some sort of sophisticated software semblance from code we’ve figured out to write.
The actual act of linking the software together is once again pretty easy. As the robots are physically combining anyway it’s a pretty simple matter to, say, incorporate USB links or even do it wirelessly. Perhaps the most sophisticated networking infrastructure on the planet is a little thing called Link 16 which is designed to allow the U.S. military to communicate with any node that’s Link 16-compatible and transfer comparatively prodigious amounts of digital information - for example, if a fighter plane detects an enemy formation that information can be spread to a wide variety of relevant allied parties instantly instead of having the pilot verbally radio it in. It’s doubtful Link 16 has enough bandwidth to accommodate an artificial intelligence, but it certainly can facilitate real-time linking and communication between artificial intelligences, and as most U.S. fighter jets have Link 16 capability you’d think a giant robot would be able to take advantage of it as well.
Or forget the linking altogether - just code the robot software to designate a specific dominant software or intelligence from one of the component robots as the control for the combined robot, or even write a new artificial intelligence that would activate when the robots are combined (and deactivate and remain dormant when disassembled back into its disparate individual components). But that’s not really fun isn’t it? It’s much more interesting to think of what would happen if individual and true artificial intelligences were to combine and then somehow interact as a single unit. Would it be a Star Trek Borg-like hive mind? Would an entirely new personality and thus a distinct and unique individual emerge?
One of those scientists is a guy named Ray Kurzweil. An early pioneer in hardware, software and digital engineering he is currently considered one of the leading experts in the development of artificial intelligence. His thoughts and predictions on artificial intelligence and how A.I. will behave have been collected into his book The Singularity is Near, which I’ve partially read (it’s a 1,000+ page tome, not exactly something you can knock out on a transcontinental flight).
As far as relevance to a giant combined robot, Kurzweil predicts that combining several artificial intelligences into a single artificial intelligence (and restoring them back into their disparate consciousnesses) is doable, and given time will be perfected to be made easy. In fact, Kurzweil predicts that artificial intelligences will be able to combine with human intelligence to form a single consciousness and back (and likewise, humans will be able to combine and separate consciousness with each other as well). The idea of combining artificial and human intelligence to form a new distinct personality/individual is something that’s been explored in various posthuman and singularity-type fiction, including one of the most well-known Japanese animated sci-fi movies out there, Ghost in the Shell, considered one of the most classic examples of A.I./singularity-specific cyberpunk.
So, at least according to Kurzweil, what would happen if two intelligences, artificial or not, combine? He’s pretty sure that the event will be tantamount to the birth of a new sentient individual - and just as you can’t predict what your own child will be like decades from now, it’s hard to predict what the traits of that individual will be. For example (and barring specific coding or software forcing a predetermined course), there’s no guarantee that a giant combined robot made up of a bunch of smaller female robots will in turn identify as female, male, or anything at all. Without any in-built coding restrictions and left entirely to the whims of free will, there’s no guarantee that Devastator would decided that maybe smashing Autobots isn’t in it for him and teams up with Optimus instead, even if his individual components still remain fiercely loyal to Megatron. There’s no easy way to predict how smart or dumb or what aptitudes such an intelligence would have, and it’s simply not correct to say that somehow it will be a mirror or aggregate of the individuals making up the whole. As a freshly birthed individual, that intelligence will simply be the product of just how things work out. The only way to be sure would be to find out for real.