Daily Archives: May 18, 2022

Tribalism, the biggest problem in engineering?

One thing that has always frustrated me is the tribal attitude in engineering best described as ‘not invented here’. When you suggest something to one of your own immediate colleagues, they are likely to pick it up, bounce it around, build on it. If someone else has an idea, who isn’t in your team, your team might fall a little behind in the competition for glory, so it is a tribal threat. The result is that many great ideas are thrown away simply because they came from the wrong people. Of course, tribalism is very multidimensional, so you might sometimes include in your team friends or distant colleagues, even people employed by competitors, and you might understandably exclude frenemies, or the annoying twat in your own team with too big a mouth and too small a brain, who just puts everything down.

For solitary workers, the problem is sometimes ego. If someone else has a good idea in your field, that could make you look less smart because you didn’t come up with it, therefore you need to find a way to shoot it down. Giving credit and praise to someone else can be difficult if ego is involved.

When a new idea is embryonic, far from its final state, it’s usually very easy to find holes in it. Sometimes of course, the holes are serious and the idea is actually rubbish. Sometimes there are engineering solutions to those holes. A good team will try to find solutions to obvious problems before dismissing an idea that might have some real value, and even if the idea is eventually discarded, there may be parts of it that can be developed or applied elsewhere. That constructive behaviour is much harder to find if you aren’t part of the team that would be responsible for carrying it through. In multidisciplinary fields, which is most things now, that kind of tribal barrier is even more of an issue because ideas will more often come from individuals with different backgrounds who are outside your team, but the standard human reaction remains tribally motivated dismissal.

A lazy dismissal technique is finding some vague similarity to a previous idea that failed. Another is to judge it by the creator, attacking the person (or department) instead of the idea. Another is to cite a problem that used to apply when technology was different, without reconsidering it with new technology.

Another is to translate the idea into a totally different one and dismiss that. I think that is the most dangerous and I still encounter it weekly. Philosophy is a common mechanism. We often hear philosophical attacks on various parts of AI for example. Taking an idea out of engineering and using philosophical jargon to only seemingly describe it allows abundant opportunities for wilful misrepresentation. It allows it to be falsely likened to other ideas with only superficial philosophical similarity and then for an argument against those superficially similar ideas to be used against it. “I can’t argue against your engineering, so I’ll drag it onto my playing field and argue against a philosophical concept I do understand and pretend it’s the same thing”.

A similar technique used whilst staying inside engineering is to simply misrepresent it, essentially deflecting attention onto something else that is more easily attacked. A common mindset may charitably be described as “If I was going to make it, I’d do it this way, and that won’t work because x, y, z, therefore your idea is rubbish”. What they really mean is “you can’t possibly be as smart as me (or the others in my team), so you probably want to do it in this obviously idiotic way, and that won’t work”. This kind of attack is amazingly common. I could put 90% of the arguments I have ever heard against machine consciousness into that category: “the brain is not a computer”; “it’s impossible to make something smarter than the engineer who writes the code”; “you can’t make something you don’t understand”. These arguments hold no water. My daughter’s brain is smarter than mine, and I have no idea how it works, so her existence is proof that it’s easy to make something smarter then you can understand. Who says your smart machine can’t be biological? As an engineer, I’m free to use anything that complies with the laws of physics.

This ‘not invented here’ tribalism probably costs the economy trillions of dollars every year, with many great and potentially valuable ideas thrown away before being properly considered. I suspect smaller engineering teams make it worse. Engineers are human, with all the faults and weaknesses that go with that. The desire for personal or team recognition, for glory, is as strong a motivator in engineering as in sport or battle. What is surprising is not that the tribalism problem exists in engineering, or that its economic consequences are large, but that it receives so little attention by teams responsible for training, team building, leadership. When there are so many compulsory courses plaguing everyday office life that often seem to address very trivial issues, how come there is so little attention to this enormous problem? Great leaders can motivate entire workforces and some companies do manage to achieve great things, but perhaps they could have done even better. Unless a problem is explicitly recognised and addressed, it’s highly unlikely that its consequences will be minimised. It certainly needs more than just an occasional team-building event.