Post
Alex Stamos
@stamos.org
1) Many underestimate the investment the PRC has put into influencing the US social media sphere.
2) The destruction of Trust and Safety at X has made it a playground for PRC actors.
3) It's nice to see them whine about BlueSky.
4) But there is nothing magical about BlueSky to keep them away.
lol China's state media is sad about Bluesky because it's not an easy propaganda distribution channel like X (and Facebook)
www.semafor.com/article/11/2...
November 26, 2024 at 11:58 PM
174 reposts
8 quotes
613 likes
From what I have seen, and the excellent Platformer from today, BlueSky is going to need some serious T&S investment and innovation to keep ahead of both prosaic safety problems and the organized state actors looking to build a foothold.
www.platformer.news/bluesky-grow...
Despite Max's snarky comment about Facebook, the Meta threat intel team has done more against organized state propaganda actors than any other company and BlueSky should start by hiring a dedicated resource to plug into this partnership (unlike X):
transparency.meta.com/metasecurity...
mentions BlueSky's unique moderation model perhaps being useful, which might be true, although unfortunately 3rd party labelers (rightfully) don't have access to the account metadata that is most useful in finding and clustering professional fake account farms.
Yes, and I worry about the quality of the feed from 3rd party labelers - the infamous "adding quad8 IP to a blocklist"-type of problem, not to mention expected bias and inaccuracies. Definitely listening to you and your insights on all the things I don't know about the problem...
zionist progressives been adding other progressives (women of color!) that oppose the gaza genocide on block lists labeled for maga and nazis. the bias is a real issue.
And nothing to stop Elon Musk or some other billionaire from buying Bluesky.
Maybe not magic , but don't you think has some novel control benefits? One big disadvantage is the lack of advertising and promoted posts. Then, there's the emerging culture of blocking and outing, which will force a shift in engagement tactics (hence further increasing cost).
Advertising has been of marginal benefit to state actors over the past several years and we haven’t seen many large advertising campaigns since the early Russian efforts. Its usefulness is overstated by those who want to blame everything on “surveillance capitalism” but is not backed empirically.
Promoted posts and artificial reach via purchased badges are really harmful affordances on X that BkueSky lacks, but there is nothing stopping any of the now veteran PRC groups from creating large organic networks from infiltrating. The BlueSky team needs to start front running and looking for this.
I do hope that the culture of not engaging and blocking early does help reduce the ability of state actors to build large audiences. Since their accounts are necessarily pseudo-anonymous they often rely upon sweeping, click-baity behaviour to build followings that hopefully the community rejects.
As this platform grows, the savvy of its median user will decline. I don't think a culture of anything is going to overcome large players exploiting the fragilities of both the platform and its users.
I agree that Facebook has done it really well, but it's taken a massive investment.
As I say above, the good thing is that BlueSky can benefit from that FB investment by hiring the right team and plugging into the threat sharing agreement with them, Microsoft, Google and others. They also need to figure out an easier verification model ASAP.
We have all underestimated how much otehr countries are prepared not just the PRC but basically all of them.
My guess is we don't do it because the cost of these influence ops is incredible.
But PRC Russia et al, have learned that we lapup Bullshit🤦♂️
Unfortunately the United States absolutely sponsors fake online accounts spreading propaganda. It’s not relatively expensive, but I do think it’s incompatible with US ideals and our best interests in stopping online manipulation by our adversaries.
cyber.fsi.stanford.edu/io/news/sio-...
It is expensive at the scale needed, and the concept of expensive, is really a function of what you see the value of what you are doing.
All Govs now do fake news/propaganda, but of course not on the scale of PRC Russia et al.
Your last sentence I can't agree more.
Nothing magical, no, but having the userbase on thousands of different algorithms instead of just one does make things harder.
One bluesky advantage:
The PRC can't buy ads to sneak into our feeds
Bluesky is a stand in for democracy, in that it requires advanced citizenship. Namely, involvement in its wellbeing by stakeholders. For bluesky that means blocks and awareness and a willingness to combat these kinds if efforts by both vigilant users and the services moderation systems.
At least I find it easier here to block and report that stuff.
Re 4): Indeed the newsletter Semafor quoted from already says so, too toosimple.substack.com/p/bluesky-ta...
All that dastardly CCP propaganda will have us believing that our "infrastructure and our institutions are collapsing". Next, thing is we'll all start thinking that a functioning state that looks after its citizens is possible! Gotta nip that shit in the bud, or soon we'll all be cursed with HSR.
Can yall dive into how an org like bluesky can handle content moderation at scale? Maybe even have or on?
Some specific topics might include:
-how will robust and accurate user verification happen at bsky?
-how will content moderation happen?
-how expensive will these efforts become?
-how will bsky monetize to grow its core team and these efforts?
-is bsky opposed to an ads biz? What else could work?
My intuition tells me they’re likely to say something along the lines of: whatever we haven’t figured out, the community will solve
The openness of this ecosystem is great but you can’t outsource every problem to 3rd party developers
It’s cool, for example, that anyone can make a feed and ostensibly the best feed can win (assuming you have access good enough data to make a great feed possible)
But you can’t outsource every unmet product need. Users will default to the most popular client atop the protocol and…
will be at the mercy of the product decisions of that app. Even if they can choose their own feeds, defaults matter and a new user doesn’t have a way to add drafts, bookmarks, group DMs, or any other missing key features
And finally, let’s assume a steel man where some third party developer does try to handle something like user verification. That will be a large, ongoing problem/effort. What will realistically pay for and incentivize one or more competent trustworthy external teams dedicated to this?
Agree -- corruption and enshitification lowered the barriers to entry, but disinformation was still a problem before algos determined what we saw.
Would the deterrent for Bluesky be our collective recognition of what happened on X and the ability to call it out this time?
Basically you can't run the same trick plays twice?
We have to be alert. I hope people who have expertise in identifying disinformation from China and Russia will advise the rest of us about how to spot (and block) these bad actors.
Whether it’s banned or not the Great Firewall is dropped to let propaganda workers through.
jenpan.com/jen_pan/50c....
People don't get it.
Be afraid. Be very afraid.
Afraid enough to do something.