Tell an algorithm to optimize for engagement and it will promote the content causing the most engagement: novelty, conspiracy theory, outrage, surprising "revelations" and all sorts of clickbait.
If we removed Twitter, Facebook and Tik Tok and reverted YouTube to the pre-algorithmic (subscription based) era, we'd still have free speech on the internet but probably a lot less bullshit.
Note: Twitter would still promote bullshit even without a ML algorithm. The rules of engagement:
- short content
- individual popularity means content popularity
- retweeting is instant, requires no critical thought
create an environment where clickbaity emotion-viruses spread at exponential rate, and things that require engaging the brain are ignored. Its essentially a crowdsourced rating system for tabloid headlines by design.
It wasn't social media who decided that the government should strait up lie to people just because it was the best way (in their opinion) to get the best outcome.
A big part of the anti-vaxx movement comes from the fact that the government have made it very clear they will lie and misrepresent data whenever it is politically advantageous for them.
If you want a very recent exemple, just look at this interview from someone very high up in the food chain at Bayer (youtube.com/watch?v=qowDwaYx7vI) and then try googling about whether or not mRNA is gene therapy. it was considered dangerous misinformation to say what he just did and yet he is overseeing the fabrication of covid vaccines.
It's far from the best example, but it's one where I won't have to waste an hour getting proper sources
It's hard to defend this without arguing for allowing the government to lie to people without any defense.
However, even if it delivers to you true information about abuses and lies being told to you, you should still be aware that the system is doing this so that you will be outraged, rather than that you will be informed.
Yeah, this is a classic example where social media spreads falsehoods, including mistakes.
Here the exec is stretching the definition of the term gene therapy to include things that don't alter genes "mRNA vaccines are an example for that cell and gene therapy". He basically makes a mistake, because the same platform is also used for gene therapy.
To use a software analogy, just because both mRNA vaccines and mRNA gene therapies are software, doesn't mean that mRNA vaccines perform system updates (instead they trip the suspicious behavior detection of the anti-virus and trigger a download for new virus definition files).
Additional subtle distortion happens later on social media when people take this mistake and "shrink" it in the opposite direction "actually mRNA vaccines are NOT vaccines because they're gene therapy".
Joe Rogan then takes that and stretches it further to imply this is why the vaccines don't last long
> This is really gene therapy. It's a different thing. It’s tricking your body into producing spike protein and making these antibodies for COVID. But it’s only good for a few months, they’re finding out now. The efficacy wanes after five or six months. I’m not saying that people shouldn’t take it. But I’m saying, you’re calling it a thing that it’s not. It’s not exactly what you’re saying it is, and you’re mandating people take it
Sinopharm makes a COVID vaccine from dead / inactivated virus. Its used widely in China as well as in the developing world. We've also seen the same reduced antibody neutralisation there, and it isn't necessarily because of the passage of time but to a large extent due to the evolution of the virus to be significantly different from the original.
So yeah, this is a perfect example of how social media spreads engaging novelty - whether its deliberate disinformation or simply mistakes - and continues to distort it further by continuously applying stretching and shrinking. The info blob continues to evolve to evade people's "immune defences" and cause them to re-share it (e.g. retweet).
This is not something new, traditional media have been doing it for decades ever since the rise of the yellow press, with varying levels of subtlety. The difference is that this here is an "organic", crowdsourced process that yields better, more convincing results, as it has to defeat the "immune systems" of many people to get a good reach via resharing. Only the highest quality misinformation makes it through.
The worst bit about this phenomenon is that corrections don't work well. People who haven't heard the rumor don't care about the correction (its not interesting) so they don't share it. People that have heard the rumor before don't like being wrong, so they don't share it either. So it doesn't matter how many scientists explain the difference, the explanation will never have anywhere close to the reach of the original (likely hundreds of millions of people)
Social media (especially short-form ones like Twitter and TikTok) is just a natural evolution of media trends that were already present in society. TV replacing radio/text and tabloid newspapers are just two examples off the top of my head. If Twitter/Facebook/TikTok weren't there, other alternatives would've been there to fill the void.
Just like we recognized tabloids for what they were, we need to recognize social media for being the same phenomenon disguised in a "grassroots" cloak (i.e. crowdsourced tabloids)
There is nothing organic about the speech that happens on YouTube. Veritasium explains the shift that happened when youtube switched to an algorithmic model from the subscriber model: https://www.youtube.com/watch?v=fHsa9DqmId8 - you need to carefully manipulate people to click your thumbnail based on the picture and the video title.
Making outrageous but believable shit up to sensationalize your content is a well known tactic of tabloids (which does continue today to some extent). YouTube and other social media incentivise exactly that.
>If Twitter/Facebook/TikTok weren't there, other alternatives would've been there to fill the void.
The particular platform that has these characteristics is arbitrary and beside the point. No one is arguing that Tik Toks particular implementation of a short form engagement optimizing feed is dangerous.
It really is just people. Look at all the information about HIV in the 80's, long before social media. People believed some truly crazy, racist, and misinformed things about HIV and its spread.
People are multifaceted beings. Social media AI automatically exploits any and all of our failings in the name of engagement. Traditional media used to do that before to some extent but the current situation is unprecedented in its reach and efficiency - custom tailored to individuals.
Technology is supposed to compensate for our failings, not exaggerate them to the extreme.
It really was crazy. Back in 1985 Neil Young apparently thought that he could catch HIV if a homosexual clerk handled his groceries (obviously nonsense and highly offensive).
"You were -too cautious- about that new disease that was killing everybody in 1985" feels more like a dark joke than a serious claim. Everybody was shocked when Rock Hudson died in 1985.
Tell an algorithm to optimize for engagement and it will promote the content causing the most engagement: novelty, conspiracy theory, outrage, surprising "revelations" and all sorts of clickbait.
If we removed Twitter, Facebook and Tik Tok and reverted YouTube to the pre-algorithmic (subscription based) era, we'd still have free speech on the internet but probably a lot less bullshit.
Note: Twitter would still promote bullshit even without a ML algorithm. The rules of engagement:
create an environment where clickbaity emotion-viruses spread at exponential rate, and things that require engaging the brain are ignored. Its essentially a crowdsourced rating system for tabloid headlines by design.