Why are synthetics in "Alien" so advanced?

Question

According to the MU-TH-UR 6000 fan wiki entry, we have the following:

The MU-TH-UR 6000, known simply as "MOTHER" was a 182 model 2.1 terabyte AI Mainframe that served as the computer mainframe for the Nostromo, Mother auto-piloted the ship while the crew were in hypersleep and monitored their activities. Mother communicated with the crew through the intercom system or a special interface room only accessible to officer at the top of the chain of command.

Compare this to its entry for synthetics:

The Synthetic's mind is an integrated Carbon 60 processor with a processing speed of 10^15 floating point operations per second. Memory capacity includes 1 terabyte of fast cache buffer RAM and 1.2 Petabyte of non-volatile memory. The system is constructed around a very powerful heuristic logic driver, making decisions based upon imported sensory data, information drawn from experience and the android's vast inbuilt databases. Intuitive functions are derived from a suite of nested contextual and semantic programs linked by self-mapping loops of tangled hierarchies.

Basically, contained within one synthetic is more processing power and ability than an entire spaceship AI.

I'm more than happy to admit that there's some inflation going on here (and the Synthetic entry isn't cited), but it has a point. In a universe of bulky mainframes and poorly synthesized speech and monochrome monitors, it seems impossible that anybody could build something as complex as a synthetic.

So, question: Was there any official explanation in the first Alien for the disparity between the Nostromo and a synthetic?

Answer 1

1. The Nostromo was old tech. Semi-falling apart, it could do what was necessary for it to do and there was no need to upgrade it. If anything, it would probably be more ruinously expensive to update the entire ship to a newer AI system when the entire ship was so old. It's the cost/benefit calculation inherent in any technology upgrade.
2. They already manufactured the synths. You don't really want to design ships with the ability to be taken over and redirected with remote commands. Far more reliable to have a solid ship that is reliable and have an advanced synth that you can just order to take the ship over. In addition the synth acts as a spy for the ship, keeps the ship in good condition, and can carry out complex orders that a ship would be incapable of.

This sounds like a bit of a reach, but I will give you some real life parallels:

1. Around 8 years ago or so, several lines in the NYC subway system went down for a few days. It turned out that a fire had broken out in the room where all the tracking information for the lines was processed, tracking where all the trains were for several lines. The technology in that room had been installed at the turn of the century- the last century. They still used century old technology to control our modern subway system. Why? Because it works, it would be an enormous undertaking to upgrade the system and still have it be able to work with the various parts that haven't been upgraded yet, and it still fundamentally works after a century.

2. A real problem that many, many companies have been dealing with in the last few years is inventory and data tracking. Why? Because back in the 60's they bought a mainframe for tens of millions of dollars and loaded everything into it. The mainframes are still used today because the company wants to get their money worth, but also because it is an enormous undertaking to transfer everything from mainframe to a modern system, get the new infrastructure in place, get everybody trained one it, etc. this is still going on today. Some companies are slowly shifting away from their ancient mainframes, but others desperately keep theirs going. And we aren't talking Ricky-dink operations, we are talking multinational corporations. That is how serious the problem is.

So investing that level of effort into the equivalent of a 1940's era dump truck would be laughable for the company to consider.