Hacker News Reader

> as soon as the algorithm begins to get what kind of content you like

I have to assume this style of algo-generated feeds wouldn't apply for an actual hiring platform, because otherwise this is ripe for "inadvertent" discrimination.

Seems like the point? Add a bit of bias laund^H^H^H^H^H^H^H^H^H^Hmachine learning and you're good to go! It's not your fault if "the algorithm" is doing the discrimination, after all.

I'm not sure how it could fly in the US where putting a picture on your résumé is more or less illegal.

Certainly would help the state plonk you down when your social score gets too low too.

> where putting a picture on your résumé is more or less illegal

How does it work for models and actors?

That’s actually a really interesting question. Models and actors usually explicitly include a headshot.

I’ve never heard that putting a picture on a resume was a legal problem for an applicant. Employers have a the burden of running a fair hiring process. Some firms even remove race and gender signifiers from resumes before the initial hiring manager screen.

It was a big deal when they stopped making hotness a requirement for airline stewards, and for most jobs it’s totally illegal to require a particular ethnicity, but for performing arts they explicitly hire people for physical appearance. What’s the difference, legally?

A few possibilities:

- performing arts (film, stage) are artistic expression, and thus protected by the first amendment. (But then why is the physical appearance of an airline steward not free speech of a brand’s image?)

- appearance is “inherent” to acting (but I’m sure people have argued that whiteness is inherent to staffing all kinds of things, and that’s not tolerated)

- actors are rarely employees, but instead sign independent contracts with short lived, project specific LLCs. Thus “casting” is not “recruiting” and actors don’t have traditional employment protections. (Most film jobs work this way though - is it permissible to apply appearance standards to the Key Grip?)

Even if we had a demographically ‘fair’ recommender, a normal recommender is super top heavy. Telling 1000 companies person x is great is not very useful to them or the rest of candidates. I would expect this system to need some heavy limiting of amount of companies a candidate sees for it to be useful. Very different than typical recommendation where a great video can be thousands or millions of views.

Ya, a perfect recommender for the "best" candidate on some qualitative set of metrics would not be discriminatory (unless the underlying metrics were).

It definitely wouldn't be useful. Hiring isn't a popularity problem as much as a matching problem.

> Telling 1000 companies person x is great

That’s the beauty of TikTok. the rich don’t necessarily get richer. You will often be presented with fresh people, AFAIK they spread out new content to measure response.

It’s not really the type of platform that everyone watches the same most popular people.

It’s refreshing that doesn’t push for “more from the same” .

I used to work at tiktok on recommendation. While we definitely do better exploration than some places there’s still a very clear top heavy bias. Part of that is just content quality. A lot of content is worse than others for any form of media. But while we are better at showing small things than say youtube, there remain very clear tiktok stars. Even for smaller areas you will still have some creators lead in that area.

I get quite a bit fresh content in the sense that it’s fresh for me(but popular otherwise) or fresh content that has a very few views and likes.

In contrast, twitter is horrible because I’m in a bubble and everyone I’m exposed to is there to push their agenda.

Instagram is fine but it’s essentially place where people I know and I don’t know pretend to be living the good life. It’s boring. The content discovery never worked for me.

Youtube is actually great in terms of content but the discovery revolves around the “more from the same” which makes discovery of something fresh pretty hard.

TikTok is not perfect but it’s the only platform that actually seems to create a stream of stuff that I like and are not the same.

So , congrats.

"The creativity tools of the app can prove very useful, the simplest idea that comes to my mind is a recruiter creating a TikTok video and says something like "Show me something that you learn in your career as X that isn't taught in school" and have the TikToker's quote and reply to the video. The recruiters then can go through the replies and pick candidates for interviewing"

This sounds like the very worst parts of LinkedIn, but with more video and far more social pressure.

I agree with you. When I'm reading resumes, I might give each one five seconds before moving on. I'd probably do the same with Tiktok applications and I'd probably have an order of magnitude fewer applications to go through in the first place.

> Show me something that you learn in your career as X that isn't > taught in school" and have the TikToker's quote and reply to > the video.

> a few second of a video can tell much more than 2 pages of written CV.

These are -different- things though, one is an interview the other is a CV.