human bias, noise, and error

Eliminating human bias, noise, and error is critical for personal endeavors and society. Mental inconsistencies plague our ability to make rational decisions and judgments over our lifespan.

Consider judicial sentencing. Research has shown that judges change their sentencing based on hunger [1]. The hungrier the judge, the harsher the prison sentence and lower the likelihood of parole.

Obviously, this behavior isn't intentional. Judges aren't consciously letting their hunger make then harsher. But it still happens.

This is where algorithms come into play, precisely because they are noiseless. An algorithm doesn't get hungry, tired, sick, or need a vacation. And therefore, those factor cannot affect it's judgement.

If you ask a computer for sentencing on the same case 100 times or a year later, you get identical results (whereas with a human, there would likely be considerable variability).

I'm not necessarily advocating for computers to take over judicial sentencing, but it is an interesting thought... Wouldn't it be more fair?

Let's consider a less charged issue: exercise. For the last few years, I've been designing my running training plan, which has helped me complete marathons and improve my personal record.

Planning my training schedule is enjoyable but tiring, especially when other things take higher priority. Sometimes, you just want the workout ready to go.

That's where Garmin Coach comes in: an algorithm that creates training plans through my Garmin smartwatch. As far as I can tell, it only considers a few key factors (sleep, heart rate, recent workouts, stress levels, etc.).

Yet, with a seemingly simple formula, it's improving my fitness and contradicting my instincts to rest or run at times I've thought otherwise.

My big takeaway? It's difficult to judge, especially when you cannot separate yourself from the data that needs judgment.

I'm the one doing the training plan, so I cannot analyze it in the third person. And I am prone to human bias, noise, and error, just like those hungry judges.

Now, an algorithm controls when I run, for how long, and at what intensity. It works surprisingly well because it lacks human bias, noise, and error.

It understands my running better than I do.

Garmin Coach reminds me of online advertising, where people often report that their smartphone must be spying on them because the advertisements accurately reflect their desires. Although I wouldn't put it past Big Tech to breach privacy, the algorithm seems to be spying on you when it's only making shockingly accurate predictions.

Isn't this scary? If you don't think so yet, you may once it gets more personal, with biometric technology built into our body, delivering real-time insights about an oncoming cold, your nutrition needs, or the questionable growth developing in your left armpit.

There are mixed emotions surrounding algorithms, especially when used as replacements for human judgment or directly wired into our skin. Many find the thought of giving human-exclusive qualities like judgment over to a machine to be dystopian and uncomfortable.

However, that might change once people see how much better algorithms are at these "human-exclusive" skills.

I think algorithm adoption is exciting, if only because they eliminate human bias, noise, and error, which may lead to more prosperity, healthcare equality, and well-being.

In 200 years, after seeing how powerfully these devices improved their lives and society, future humans may look back on our reactions to this coming wave as overblown or unwarranted—shortsighted beliefs that couldn't foresee the imminent utopia. Or maybe they'll say, "We should have listened."

I sure hope it's the former.

[1] https://en.wikipedia.org/wiki/Hungry_judge_effect