The trouble with your understanding is that you probably know very little (or nothing) about the ways and means of self-teaching and self-modifying systems. Also that you take the stipulated scenario “verbatim” and not just as an illustration of the evolutionary process.
Yes, there is a starting point, which was programmed by the programmer. (Just like we have a starting set of information processing. Nature, versus nurture.) Then the program starts to modify itself, randomly. Most of the attempts are useless, and in that case another path is chosen. The point is that even the originally “programmed” aims can be overridden and the final product does not resemble the original at all. It will choose new aims and new methods to reach them. There is no significant difference between human learning and the machine learning. (tell that to scientists studying AI) Both are processing new data, incorporating the new data into the system, modifying the processing algorithm. Forget the Asimovian “three laws of robotics”. Let your imagination roam. That is where the fun begins.
This is all about learning process, and nothing more. A computer will be able to gather and use data, but this “using” is limited.
The data collected is also limited, but for the sake of argument, let’t pretend a machine is able to have every information I currently have. Chances are, it will crash, as it doesn’t know exactly what it is supposed to be doing (my current situation

).
Best case scenario, it will try to take over the world, based on some of my “wants”. It will study hard, pass entrance exams to great Universities, learn German and be successful at life. What it won’t do is laze around, like me with my silly free will.
Still, no freedom for Machine-Me.
No, seriously, ask yourself: what would a machine do if it had enough information to do it? What would a machine want?
You missed the point. The libertarian definition of free will has three parts.
- There is an aim that we want to achieve.
- There are at least two ways to achieve it.
- The locus of decision is with the agent.
Observe, it says nothing about the internal structure of the agent. HOW the agent reaches the decision is irrelevant. What matters is that it is HIS decision, not forced or imposed from the outside.
So I was still on point. How Winston reached his decision was irrelevant (and I didn’t even dwell on it, why did you bring this up?), but the decision was his, and not the Party’s.
Again, differently from the car: it can’t make a decision that was not programmed into it, directly or indirectly.
Now this brings up an interesting problem. How free are we, humans from our upbringing (programming)? During our developing years we learn certain behavioral methods. For example: “do not torture animals”. If these methods are very deeply ingrained, we would not give in to some urge and would not torture animals. We certainly COULD, but we do not want to do it. Does this man that we are “not free” to choose “not to torture”? Just think about it.
How free are we: look at animals and then look at us. We all have the same basic programming - we feel hungry, thirsty, lust, tiredness.
Take tiredness.
An animal, when tired, will sleep. If it cannot find a good/safe place to sleep it will look for it. It will keep looking for it unless something more urgent (say, hunger, thirst, potty time) shows up, and whenever a good place is found, it will rest. It follows the program.
Humans, however, don’t work that way. Proof of that is the many teens right now who, even though they are tired, would rather keep on playing their MMO games (or watching Netflix) instead of sleeping. Sometimes, they keep on playing instead of peeing (and end up soiling themselves).
Though I’ll admit: we are looking more and more like animals nowadays, with pleasure as the main reason of our every action. Virtues are devalued; anyone trying to live up to any virtue is judged as some sort of deranged fool.
Virtues are usually the proof of our free will. Even though we can eat more, we stop ourselves when sated. Even though we can sleep more, we stop ourselves when rested. Even though we can hoard more, we donate what little we have.
We are capable of going against the “program” if we so want. You just need to try.
You bring up an interesting point, in that our development has a say in it. From what I got from it (I actually had a course on this in Nurse school), it is not really a programming, but a learned behavior.
“If you do wrong, you are punished”, so obviously most of us will
choose not to do wrong. And yet, there are the little devils (guilt here, I was a pest) who like to test the rules and often go against them. I didn’t kill animals, but I did bite down on my cousin’s finger until I heard a crunch. I used my free will for evil, and as bad as that is, it is proof that I had free will.