Chapter 3208: Product Obsessed
It became clear that Maikel had not spent much thought on what he was doing. Ves shook his head in disappointment.
"Why do we exist? What is the purpose of mech designers?"
"Mech designers exist to serve mech pilots." Maikel parrotted like a dutiful student.
"Then why aren't you trying to do that?"
"I am thinking about it! I just need to study as much relevant knowledge as possible before I can start properly!"
"You're not thinking in the direction that I want you to think." Ves frowned and leaned forward. "You know the saying but are you truly following this creed or are you just paying lip service to it? The reason why mech designers don't like to delve too much into AI systems is because much of it goes against the spirit of what we aim to do. While it is a good idea to automate a lot of small and minute functions for something as big as temperature regulation to something as small as performing voltage micro adjustments onto a processor chip, all of it is aimed at reducing the operational burden on the mech pilot. No human can control millions of different little parameters at a time."
"I'm also trying to do that, teacher! The mech always needs to help out with controlling all of its systems. What I intend to realize is exactly intended to help mech pilots fight better! Two minds are better than one, and if the mech notices an incoming threat, I can allow it to react before the mech pilot becomes aware. This crucial difference of several milliseconds to several seconds of reaction time can mean the difference between life and death!"
Ves let out a deep breath and pressed his fingers against his forehead.
"I get what you are saying. Your intentions are good, but have you ever thought about the wider implications of your chosen direction? For example, in the scenario you've described, what if the mech wants to steer itself out of danger, but the mech pilot insists on staying its course in order to take out a strategic objective or defeat a threatening opponent? Who will take primacy in this case?"
"Uhm.. I would program an algorithm that would dynamically weigh the cost and benefit of intervening."
"So you are willing to let a machine determine whether it is worth it for a human to go through with his decision to sacrifice his life for the greater good? How far will you set the threshold? If it is too low, your system is pointless. If it is too high, then there will be too many cases where the mech hijacked control when it clearly isn't desirable."
"I would only allow the mech to take action if the mech pilot is in agreement!"
"Agreement requires thought! You just painted a scenario where every millisecond of reaction time is critical. If you wait for the mech pilot to make up his mind, then the time advantage is almost entirely negated. The pilot would have been better off taking action himself!"
There were so many dilemmas and other problems associated with allowing mechs to be controlled by active AIs that it was rightfully shunned in the mech industry. Mech pilots should never be put at the mercy of heartless, emotionless algorithms that made life-and-death decisions based on cold logic.
Maikel looked troubled, which meant he hadn't put sufficient thought on these deep and difficult dilemmas. Perhaps he was already aware of them but didn't think he needed to answer them at this early stage.
"You need to understand the context behind AIs and their wider implications surrounding mechs before you can go any further." Ves advised. "It will do you no good to start designing mechs with increased automation if you aren't even aware of what it will mean to your own clients and customers. I'm not telling that your design philosophy is wrong, but you need to apply it in a way that falls in line with our purpose. Do you recognize your mistake?"
"I.. I think I do." The younger Larkinson replied. "I've been thinking too much about making my mechs stronger and more effective in battle without taking the pilot into account."
"You've been focusing on the machine in isolation, yes. Your approach also assumes that mech pilots are either incompetent or makes mistakes that need to be compensated by their intelligent mechs. While I don't object to the idea of having an AI on hand that can help the pilot survive, what do you think will happen if this approach is adopted on a wider scale? Imagine a time where you have managed to realize your design philosophy and popularized your design philosophy. Would mech pilots be better off in this hypothetical future?"
Maikel thought deeply about this. While he had fantasized about it before, Ves already hinted at the negative consequences of his path.
"The total package should be stronger." He slowly said. "That's what I'm aiming for. While the mech is unquestionably stronger, I'm not sure whether the mech pilot also benefits."
"Do you know what I think?" Ves tapped his finger against the worktable. "All of this hand holding will ultimately make mech pilots weaker. This is because the kind of AIs you seek you add to your mech designs are mostly reactive. They function as a safety net for mech pilots. If the latter falls, the AI will always be there to catch them. In the short term, this might indeed save a lot of lives, but what about the long term?"
"I think… if mech pilots don't suffer from the consequences of their mistakes, they will keep doing." Maikel reluctantly admitted. "If younger mech pilots see older mech pilots getting sloppy only for their mechs to bail them out, then the next generation won't spend as much effort to train the skills that are relevant to these incidents."
"You finally get it." Ves smiled. "You can argue whether the whole point of mechs and its suitability as a weapon platform is still appropriate or not. However, once you commit to becoming a mech designer, you must become married to the concept. It's okay to tinker with it and put your own spin on the entire idea. That is what I am doing by trying to make my mechs and live and changing the relationship between mechs and mech pilots. However, I have never attempted to undermine the fundamental principle that mech pilots should always be central and that humans must always remain in control over their own technology. What you need to do is to find a way to reconcile your goals with the same principles."
"I see… I guess I need to rethink my entire approach." Maikel said with an embarrassed expression.
Ves reached out and patted the younger Larkinson's shoulder. "Hey, it's alright. You've made a mistake that every mech designer and engineer makes. It's quite fascinating to get embroiled in all of the science and tech. Harnessing them in the right way opens up a lot of possibilities. The danger is that you can become too embroiled in your own little world and forget that products must also provide value to its target audience."
Designing mechs was a job. Mech designers earned their living by being useful to society in some way. Those who missed this basic point simply weren't qualified to advance to Journeyman.
If Maikel was just a random assistant mech designer, then Ves would have just let his student learn this lesson on his own or not at all. However, the mech design student could potentially offer a lot of support in the future, so it was worth spending some time to nudge him in the right direction.
Of course, Ves was also aware that his attempt to handhold Maikel's development might also cause the aspiring mech designer to develop a dependence on external guidance.
This was the tricky part about teaching students. There were some school systems that only focused on results and provided an excessive amount of guidance and instructions to their pupils.
This resulted in underprepared graduates who entered the workforce without even having learned how to wipe their own butts! They constantly needed to ask their managers and supervisors to do the wiping in their stead!
Ves did not want to turn Maikel and Zanthar into dependent chicks who constantly asked him for help at every opportunity. They needed to learn how to solve their own problems and find the motivation to push through the more difficult, tedious and unpleasant aspects of their profession.
As long as they were able to do that as Ketis had done, they would surely be able to come into their own. He still held high hopes for them despite his various concerns.
Once Ves finished giving Maikel some more in-depth guidance on a couple of technical subjects, he left the student be and swept his gaze through the rest of the design lab.
Around 150 assistant mech designers were quietly working on a lot of miscellaneous tasks related to the Chimera Project and the Bulwark Project.
The Design Department saved the best and most difficult expert mech design projects for last.
The Chimera Project was already rather complicated due being based around a hero mech frame. The addition of mounted wargear almost doubled the workload required to design the entire package.
The Bulwark Project on the other hand was a single, hefty expert mech that naturally took a lot more time to design and optimize. The headaches surrounding this project rose sharply when the mech designers had to take into account that it was meant to be an upgrade to an existing, inferior mech.
Ves wasn't sure whether it would take three or four months to finalize both projects despite all of the substantial progress accomplished in the previous months. The two were truly a lot more demanding than the previous four expert mech design projects.
"It's worth it, though."
The amount of hours spent on designing a mech was not an indication of its performance. However, he could already imagine either of them outshining the rest due to accommodating more mass and features than the simpler and skinnier expert mechs.
"Of course, their design budgets are a lot more extravagant as well."
Ves didn't even want to look up the projected costs for both designs. Even when he excluded the estimate on the monetary value of Unending alloy, the mech designers had already stuffed a huge amount of expensive, high-quality modules in the projected designs.
Would the added cost result in a proportionate increase in performance of the two mechs? Most definitely not. Was it still worth it? Yes!
"If we leave out the masterwork variable, then it is highly likely that either of the two remaining projects will become the crown jewel of this design round."
The Dark Zephyr was a fine expert light skirmisher, but its limited size did not give the mech designers much room to express their creativity.
The Amaranto definitely exceeded everyone's expectations. As long as it remained as the only masterwork expert mech of the current batch, then it may very well be able to hold its crown.
The just-finished Vanguard Project was a serviceable expert mech that certainly had its place in the Larkinson Clan's mech lineup, but it showed few signs of brilliance.
"The only other expert mech that can surpass the rest is the Decapitator Project."
The expert mech design was shrouded in a lot of uncertainty. Ves didn't understand Ketis' design philosophy and was unable to determine how powerful her contributions would be. He didn't have any good reference material to base his estimates as this was her first proper swordsman mech design.
He also wasn't able to determine the consequences of holding a massive ritual while fabricating the Decapitator Project. It was worth a try to see if it helped in creating a better expert mech, but Ves was not sure whether it would produce the intended effect.
"Who knows what will happen if we bring a lot of enthusiastic Swordmaidens and Heavensworders together."