Priming is one of the least reproducible effects out there and lies literally at the core of the reproducibility crisis [1]. I'm going to call BS on claims of some sort of spooky weaponization of it.

Edit: To add necessary context for those unaware, TikTok's audience in the US is politically liberal. The buy out of Trump rally tickets a few months ago was organized there [2].

It's true that we should be talking about the security implications of foreign control over our social media platforms. It's also true that we should be concerned about a targeted ban (instead of, say, a general crackdown on foreign social media) on what amounts to a meetinghouse for Trump's critics by the Trump administration itself.

[1] https://replicationindex.com/2017/02/02/reconstruction-of-a-...

[2] https://www.forbes.com/sites/jemimamcevoy/2020/07/07/tiktok-...

To quote the link of yours: "Five years later, Kahneman’s concerns have been largely confirmed. Major studies in social priming research have failed to replicate and the replicability of results in social psychology is estimated to be only 25%"

I cannot discern whether author says that social priming research has replicability of 25% or social psychology has replication success of 25% and priming research has even less.

The scale at which TikTok operates is much larger than any studies that can be performed by a group of scientists. This is true for audience (or test subjects), amount of hardware used and for information gathered.

The link you provided still allows for some replicable effects to be present. Effects that were hypothesized, planned for discovery and then tested and brought to the light of day.

At the scale of TikTok, effects can be discovered without hypothesis first, just by using sheer analytical power available.

The quote you pulled spells it out:

- Social priming does not replicate;

- Social psychology has a replicability rate of 25%.

Let me offer you other quote from different link [1] (Nature): "Equipped with more-rigorous statistical methods, researchers are finding that social-priming effects do exist, but seem to vary between people and are smaller than first thought, Papies says."

[1] https://www.nature.com/articles/d41586-019-03755-2

Effects do exist.

> but seem to vary between people and are smaller than first thought, Papies says. She and others think that social priming might survive as a set of more modest, yet more rigorous, findings.

You linked the full statement, and then focused in on only two words of that statement for your comment.

What about the linked article makes us think that this is something that the US government has the authority to step in and ban? We should skirt both the 1st Amendment and the free market over a social theory that might survive as a set of modest findings?

It's good that the courts are blocking this in the absence of a specific, credible threat. "Some effects do exist, but we don't really know what they are, and they're a lot more modest than the scare effects that people are familiar with" is not a specific, credible threat.

The threat of allowing the US government to get comfortable exercising ever more invasive control over the free market for questionable reasons, to the point of turning off or banning widely used communication channels -- that's a much more tangible threat than what I'm seeing linked in your article.

I am not talking about dangers of US government banning some app or service. I am talking about about dangers of the very existence of that service ruled by what I consider quite unfriendly people. Unfriendly not to me personally, but to the way US operates.

As usual, official justification and real reasons may be very different. The "security threat" may be direct and indirect. I tried to present a picture of indirect threat to the security of US.

> I am talking about about dangers of the very existence of that service

My point is that the indirect threat to the security of the US through priming is statistically tenuous, that there's little reproducible evidence even from optimistic researchers that priming works on a significant scale (or at all), and that even if the threat is real it's likely much smaller than the threat posed by this kind of over-regulation and government overreach. My point is that a threat this vague doesn't warrant this kind of discussion in the first place.

The link you posted optimistically describes a very limited, focused effect that is still actively being debated. Is there any accepted scientific evidence, at all, that priming in a 60 second TikTok video would have a larger effect on the average person's politics than seeing friends share fake news on Facebook?

I would like to note that your last question equalizes single 60 second video and fake news post. You are discarding effects of repeated viewing of different algorithmically selected videos on the views of viewers. It is an experiment that is being done right now by no less than TikTok itself.

Actually, my beef with the Facebook is exactly the same - one does not control the feed, the feed is being controlled by Facebook.

> You are discarding effects of repeated viewing of different algorithmically selected videos on the views of viewers. It is an experiment that is being done right now by no less than TikTok itself.

And is there any accepted scientific evidence, at all, that this would change whether we should be more worried about priming effects in a TikTok video than fake news posts directly shared by friends and family?

> Actually, my beef with the Facebook is exactly the same - one does not control the feed, the feed is being controlled by Facebook.

Our concerns with Facebook aren't related to priming. If your concern is that centralization and control over algorithms can be dangerous (especially in the hands of an authoritarian regime like China), then I agree, but I don't see any particular reason why TikTok should pose a unique danger in that regard. CraigJPerry's original comment you replied to still seems pretty on-point:

> Surely Facebook’s a bigger issue than 60 second videos since any bad actor can target messaging at susceptibile users and no one will know.

> [...] what is the method of using a 60 second video platform to share messaging with a crowd who will immediately scroll past anything heavy (or just plain stop using the app) to get to the next funny video? It seems a ultra low effectiveness medium for controlled messaging.

The main conclusion I have is that TikTok may pose a danger (just like any social media network), but the dangers it poses are not big enough or well-defined enough to justify this kind of intrusion into the free market. Is that something you agree with?

You keep asking for scientific evidence of effects of completely new phenomena. I cannot provide you that. On the other hand, I am pretty sure you cannot provide me with evidence of absence of effects of the experiment of that magnitude.

Facebook is also doing priming, by controlling what user sees. Heck, even television and radio back in the day were used for priming, albeit not in such a direct feedback way.

About "intrusion into the free market".

I see "free market" as a neutral or even negative thing. I do not see "intrusion into the free market" as necessarily negative thing.

I see ban of TikTok as somewhat positive thing given my view on situation. I really want to see regulatory actions on Facebook too.

> I am pretty sure you cannot provide me with evidence of absence of effects of the experiment of that magnitude.

This is not how regulatory policy should work. I'm just going to repeat my first comment:

> It's good that the courts are blocking this in the absence of a specific, credible threat.

> TikTok's audience in the US is politically liberal

If anything, then that audience is a very interesting target for either:

- trying to convert them to more conservative ideas, by feeding them with messages that play on their feelings

- trying to get them to not vote, by discouraging it

- trying to incite violence against the other group in order to destabilize the country

Especially the last one is interesting for China. The second one has been (by their own word) effectively used by Cambridge Analytica in Trinidad to influence an election by discouraging young people to vote [0]. Imagine if there was an app, mostly used by impressionable teens and young adults, and you had unlimited access to the users of that app.

[0] https://www.youtube.com/watch?v=omc-5zj70M0

> Facebook is primarily text which is low bandwidth and it goes through higher thinking.

Today Facebook is mostly images, livestreams and videos.

My wife uses Facebook and I beg to disagree with you.

I use Facebook: my experience is it's probably 50-50. Which means far more content (by duration or bandwidth) is images/video. I rarely look at the videos.

As an erstwhile small business owner it became very clear that business page posts with images always get more views so all informational posts always have images if you want to improve your engagement scores.

Most friend posts that are popular enough to spread to me include a gallery or shared video.

People make text comments on news stories and link a news website, which then pulls in an image/video. So often the headlines/images are misleading - as one finds on HN too.

I used TikTok for 6 months or so, even posted a video. To me the only content was ever goofy dances, magic tricks, acrobatics and 'pranks'. No more harmful than any other human-targeting Skinner Box.

Does your wife heavily use tiktok? It's possible that different cohorts of Facebook users experience the product in different ways. I can imagine that those drawn to Tiktok also make use of insta stories or fb stories (short videos) etc.

Your feed is customized to your interests. Some people get lots of words, others just get images, livestreams, and videos.

> Facebook is primarily text which is low bandwidth and it goes through higher thinking. TikTok is high bandwidth from the start and is primarily analyzed using much older parts of brain.

Agreed. Reading and writing are recent inventions and involve thought and deliberation. There's a reason why such skills were rare and highly valued in antiquity. It's reasonable to assume that a medium with an emphasis on text is less prone to (but not entirely exempt from) manipulation.

On a somewhat unrelated note, people like Elon Musk believe that "information overload" can be solved by increasing the bandwidth between the medium and the brain. I wonder if that's true, or whether the lower bandwidth is necessary because our newer brain (prefrontal cortex) simply isn't fast enough.