Does everyone agree AI is better than self discovery? | Hacker News Reader

Does everyone agree AI is better than self discovery?

1 pointmanabovethesky4 comments

I get that computers can do things quickly, but this massive push into AI and pre-selected candidates seems somewhat misguided. Perhaps for high level areas, but not for something more meaningful or deep? Currently, I'm building this site https://nextcept.com/Feed/Network to help people find others to work with that they get along with. Ie. cofounders, early teams, new hires, etc. I continue to get suggestions that I need to offer suggested or recommended candidates. I suppose I see this to some degree, but am I alone in wanting to be able to wander through candidates, learning about them individually, until I find someone interesting. Obviously we're trying to make this wandering process more efficient, but it seems like something unique or something that stands out would be difficult to capture in an algorithm. Is this a misguided belief? Does everyone want (believe it's possible) the ease of Tinder but the depth of a real connection - magically?

Lecturer and teacher Scott McCloud once said this to the effect of hiring practices, how there are three anchors from which an employer of any variety must seek out among potential employees.

To turn in work of professional and skilled caliber, to do so always ahead of deadlines, and to be easy to deal with generally.

A person only offering one is not providing enough to make their employ worthwhile or long-lasting, while the chances of someone realistically offering all three are extremely rare. So, the idea is to look for somebody what can comfortably guarantee any two of the three anchors. For example, a person may only submit average-quality work, but if they are reliable for at least that much and are easy to work with, then they are employable. Or, a person might be notorious for blowing schedules, but if they are always friendly about it, and their end product is always outstanding even though late, then they themselves are worthwhile.

But no psych test in the world can guarantee with 100% accuracy exactly how people might work together. Too many variables and too many subjective variables.

Not magically.

I used to put down the leaders of big businesses for the various screw-ups you read in the news, such as the 2008 financial crisis, Carly Fioria destroying Bell Labs and nearly destroying HP, the recent meltdown at GE, the 737 MAX scandal, etc.

Over time I realized that incredible screw-ups happen all the time at small businesses (say a small-town general store) or medium businesses (say a venture-funded startup) If the people who ran big companies screwed up as often as the people who run big companies, we wouldn't have any big companies!

You might not need a 10x salesman to be a 10x developer, but the failure of any part of the value chain can multiply the value of your work by zero!

There probably is some way to measure the pathology that I mention, but it's not going to be superficial or "magic".

Most people in dating have deal breakers. I feel like there should be some way to filter out deal breaker issues.

This is a good idea. I've been trying to build in methods for determining if you'll get along with a person, but the idea that you might be able to filter based on deal breakers is interesting. Or to filter based on what's not needed. I think there's this motivation for companies no a days to aim for their perfect candidates who know 40 languages and I never really know why. Is that really their goal? Do they just want someone to say they know all of those things? Why not just focus on what's needed and what's needed a little down the road. It seems like if a person could help with that, they'd be helpful in the future.

Reply on news.ycombinator.com