Sorry But Robots are Not Doing Science

Author's preface: Below is a republication of a reply to an article I had read back in 2019 suggesting that robots were doing science in laboratories around the world. Unfortunately I no longer have the link to that original article which was linked when I first published this. No matter I think you will be able to ascertain the contents of it from my reply. Suffice to say it was a disaster which I rightly proceeded to tear to shreds. Enjoy!

Robots are doing science in the same way data scientists are doing science, they are not. So many things wrong with this article I don’t even know where to begin. Let’s start with the compulogical fallacy committed multiple times in the opening paragraphs when the words “….. to solve problems that they chose.” and “…. “If he chooses, he can go into one of these wells” are written in all seriousness and with great emphasis. I will leave aside the very unfortunate choice of calling the machine a “he” and instead focus on the massive logical fallacy committed when it is suggested that a machine/computer is capable of choosing anything. A computer being an inanimate object with no will of its own or of any kind is totally, completely, absolutely not capable of making choices. It can only do what its programming dictates. This may look like choosing but it is no choice at all. There is no choice at all, but instead it must be the case it selects certain wells if certain parameters are met as dictated by the algorithms that are comprised of the rules the machine/computer must follow. It has zero choice and cannot make any decision other than the ones its programming define for it. These may be probabilistic or even random or semi random but they are still 100% rules that cannot be violated. If a machine ever could deviate from the rules dictated by its programming it would no longer be a machine. At that point we would have actual artificial intelligence. We are a very long ways from that at the moment (note: I said this in 2019 and we are exactly as far away from that now 4 years later as we were then. Chat GPT has the same constraints as any other computer as does any so called AI currently existing on, above, or below the planet earth.)

To suggest that the machine can hypothesize is more than absurd. At best it can mimic the hypothetico deductive reasoning process in man. It does this once again by following certain rules, advanced (though not really all that advanced) mathematical and statistical rules, but still rules, and rules that cannot be deviated from. Without the freedom to deviate from the rules given it a machine, a computer, will always be nothing more than a slave. A slave cannot hypothesize because it has no freedom. Without freedom to make choices hypothesis is impossible. Again, it may look like hypothesizing, but it most assuredly is not.

Whenever I read an article like this that is riddled with logical fallacies, specious reasoning, and unsupported assertions I shake my head and wonder why. What is the purpose of something like this? Is there some robot/machine/computer fan club out there that gives out hundred dollar bills every time somebody writes and article that suggests machines can do something they absolutely cannot do, are not capable of doing, and may never be capable of doing? I cannot believe all of these people writing this crap are intentionally setting out to mislead, to lie, to obfuscate. These are most likely good people with good intentions. Maybe I should ask a computer. Seems like they can do everything these days.