OpenAI is very smug after thwarting five ineffective AI covert influence ops OpenAI on Thursday said it has disrupted five covert influence operations that were attempting to use its AI services to manipulate public opinion and elections. These influence operations (IOs), the super lab said, did not have a significant effect on audience engagement or in amplifying the reach of the manipulative messages …
Yep! AI efficiently automates it though. It's like fast junk food for the mind, unhealthy and highly addictive, and now produced by robots, 24/7, right on time to sway the EU elections (June 6-9, 2024) into surrendering all resistance and weaponry to the interests of the International Union of Virtual Bad Grammar STOIC Spamouflage Media.
I'd suggest a quick check on the 38 parties (or voting blocks) to see which of them take positions that support the usual automated-propaganda suspects (Russia, China, Iran), and voting at the "opposing end" of that ... or just voting Pirate Party when in doubt (whiskey!).
That said, the CETaS report argues that AI content creates second-order risks, such as sowing distrust and inciting hate, that are difficult to measure and have uncertain consequences.
So, no different then from the likes of international television organisations and printed word operations which have been hosting and posting news on terror campaigns since forever they were conceived and perceived to be capable of and able to lead both localised and globalised public opinion in a certain direction .... remotely and autonomously/virtually invisibly and relatively anonymously.
And ..... even though you may dislike it, to deny it surely identifies you as ....... :-) well, a combative non-combatant would be one possible sad AI moniker to pretend was not true and decline to acknowledge is useful for serial abuse and constant misuse creating a false state of present normality on which to build a future of further virtual realities/Advanced IntelAIgent BroadBandCasting Scenarios ‽ .
Bravo, AIBBC, and thanks for the Mined Minds and Greater IntelAIgent Games in Play for Work, REST and Play.
"We all expected bad actors to use LLMs to boost their covert influence campaigns — none of us expected the first exposed AI-powered disinformation attempts to be this weak and ineffective," observed Thomas Rid, professor of strategic studies and founding director of the Alperovitch Institute for Cybersecurity Studies at Johns Hopkins University’s School of Advanced International Studies
And therein is the problem, the first exposed. Then there's all the ones we don't know about.
Having said that, looking at the online discussion of the Orange Convict's latest newsflow, or online discussion of events in Gaza, I have to question whether there is any difference between irrational, biased shouty garbage from maybe-sentients and the same irrational, biased shouty garbage from non-sentients.
Yep, it does sound as if these are the very lowest-hanging fruit of their "operation". Or possibly just windfalls.
It would be more reassuring if they'd discovered at least one reasonably sophisticated operation.
But here's a prediction: they will. Shortly after 5 November this year, we'll hear about a sophisticated operation favouring whoever won the US election. If Trump wins, there will be a lot of talk but no evidence. If Biden wins, there will be plenty of evidence but almost no talk. Either way, Americans' faith in their election process will take another poke in the eye.
And therein is the problem, the first exposed. Then there's all the ones we don't know about. ..... Lurko
Some folk will tell you that there is nothing new available under the sun, it is just a case of you having not recognised or been exposed before to that which may be of growing concern to those with every reason to be concerned.
It is quite amazing just how quickly so many appear to forget or know so very little about anything worth knowing anything and/or practically everything about.
Reports that say that something hasn't happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns—the ones we don't know we don't know. And if one looks throughout the history of our country and other free countries, it is the latter category that tends to be the difficult ones. ...... Donald Rumsfeld, United States Secretary of Defense, February 12, 2002
And that makes SCADA Administrative Systems catastrophically vulnerable to external remote exploitation both for and with alien commands in almighty overall virtual control.
How far down the slippery slope have we slipped that Donald R. (arguably a blood thirsty, geriatric, avaricious, retarded ex-spook) appears the model of rationality and sanity when compared with the late US President Donald T.
Rumsfeld was basically avoiding answering a serious question by muddying the waters with this nonsense.
He might as well have said "you have a colouring book with pages that you have coloured in, some partially, others not at all, but also contains other pages that are completely blank."
He could have more succinctly stated "at any time there are always things [threats] to which we are completely oblivious. Historically these have been the most challenging to both the US and the West generally."
looking at the online discussion of the Orange Convict's latest newsflow, or online discussion of events in Gaza, I have to question whether there is any difference between irrational, biased shouty garbage from maybe-sentients..
In fact those of us of solipsistic persuasion are just eagerly awaiting the enhanced intellectual level of online debate AI will bring. Bliss
I would expect any decent sized operations that want to do anything dubious with "AI" to run their own, plenty of lower parameter number LLMs you can run on a decent spec machine that will do a less good (but likely "good enough" job) so keeping more under the radar than if you use a third party providers public offerings (or sign up to their API)
Though given how poor "AI" output often currently is, probably better getting actual people to write stuff, unless you need to churn out a lot of varying content, until "AI" gets a lot better.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
