I think more drama has been created around this than is necessary. Based on the video, the real-time projected visitor's faces were not analyzed. They were simply shown with a random description flag attached, such as "energetic," "compassionate," "inspiring," "fitness influencer," or "cloud watcher." It seems to be an artistic provocation showing what a real people analysis could look like.
But now I can point a camera at a crowd and It will:
- count the number of people and animals there
- give me an estimated gender for each
- analyse the sentiment of each person
- save their facial features so I can find "Male-sg76fg" in future photos automatically
- store the GPS location
With some extra work I can then find each of them on social media, grab their real names and other information from public sources and now I have a surveillance database. (Illegal where I live, but who's gonna check?)
This makes "public photography" a whole different thing from what it used to be.
IDK about shouldn't. Public photography not being a crime comes from a time where one could still be generally expected to remain anonymous despite being photographed. Just like how you can be seen by strangers in the street while walking and still remain anonymous. Yet stalking is a crime, and facial recognition seems to be the digital equivalent. Facial recognition is something that can be done at any point by someone with your picture in their hand.
This means you can not make a photo/video of a person in public without their consent if they are the focus of your image. They also have the right to revoke consent anytime in the future.
The only exception is at large gatherings like for example the Street Parade where the expectation of privacy can not be expected especially since the event is televised.
This is also why you can not put cameras on your home that film public streets etc. They need to be blocked off or facing the other way.
IIRC some countries recently started experimenting with automagically granting copyright to people for their own likeness, I think it was aimed at AI generates fakes, but it's probably more widely applicable.
Anyway, don't be a dick, don't take pictures of people without their consent.
See https://www.photrio.com/forum/threads/law-regarding-photogra...
Where I live, a concert is not considered "public", unless it's a government-run event on government property.
Otherwise, a concert is a private event, in which case you have no right to privacy. Just like going into a store.
Or
"Data shows you hang out in low income areas, we don't think that aligns with our companies goals."
So the "face your principals" is completely fucking arbitrary. That's the fear.
It’s incredibly common for tickets to big gigs to have fine print along the lines of “by attending you consent to being recorded”. This has been the case for decades. If you’ve ever watched an official recording of a live performance, you’ve seen this in action.
This is just a novel presentation of what is already commonplace recording. And it’s great and it makes a point, but the article is bad.
I that case they should have used descriptions like "gay", "muslim", "poor", "bipolar", "twice divorced", "low quality hire", "easy to scam", "both parents dead", "rude to staff", "convicted felon", "not sexually active", "takes Metformin", "spends > $60 on alcohol a month", "dishonest", etc.
None of the people who actually take advantage of you or manipulate you using surveillance capitalism cares if you're a "cloud watcher" or "inspiring"
That would certainly better demonstrate the scary dimension of mass video surveillance and face recognition. However, not many people would buy tickets for the next Massive Attack show after being lectured like this.
One season Saturday Night Live did this with its studio audience as a recurring gag.
The one that stuck with me was the couple labeled "Pregnant two hours."
Their images were not being sold, nor were they being used to promote the concert. Plus, nearly everyone who goes to a concert these days agrees that their image will be captured and possibly used in future promotional material.
Also consider resourcing, the manpower, money, tools, electricity devoted to surveillance back then compared to today
How about today? Where could you venture in secret without being tracked? How could you hold a private conversation? Your face & license plates are constantly tracked, along with your personal phone, laptop , watch, fitness tracker, Tire Pressure Management Systems, etc.
If you had to assign a logarithmic authoritarian intensity scale to those regimes, and to today's regimes, how would you rank them? Consider the spying capacity, resources, recording capacity, analytic capacity.
I would put today's regimes many orders of magnitude more severe.
what do you think?
My head hurts.
[1] https://news.met.police.uk/news/arrest-landmark-for-met-offi...
[2] https://www.bbc.co.uk/news/articles/c62lq580696o
[3] https://www.independent.co.uk/tv/news/met-police-facial-reco...
Authoritarian regimes come up with bogus charges to include political opponents in the "bad guys", painting them as criminals to the rest of the country, and legitimizing their arrests.
These surveillance technologies have two main problems: if you have more data, it's easier to dig dirt on people. And if you don't have data, you can always fake it.
The requirement for it to work though is that you need regular people to believe that political opponents are in fact criminals.
The scary shit is that the US is not too far from being there.
In the past decade we had some examples where some countries had really big anti-government movements and protests... and they simply couldn't achieve anything substantial. Iran comes to mind, but I think we also had some in the East of Europe.
And then there are countries like Russia and North Korea (and likely many more) where it looks like (at least from the outside) that mass protests are pretty unlikely, because any kind of political opposition is suppressed before it reaches this level.
Also the martial forces (police , military, security ) are more directly managed , and more broadly deployed . You can no longer reason with an individual because their decisions have to be run up the chain . Individuals no longer have authority to provide exceptions or help
We can imagine something like 1984, where it was only the party (middle) class monitored for thinking. But proletariat (low) class was free to think whatever - because they knew system was bad; they weren't required to pretend.
I guess my point is, the totalitarian system doesn't need widespread surveillance. But it needs believable ideology, which enough people from low and middle class believes to keep the communication barrier between these groups sufficiently closed.
Is the neoliberalism such ideology? Is it something that can offer enough positives to sustain/counterpoint negatives of widespread surveillance? I doubt it.
We can look at recent examples where UK and US tried to control the narrative and failed - Palestine Action, ICE arrests, troop deployments in cities, Tiktok ban.. Despite surveillance, people are not buying the ideology.
I do, however, also appreciate how strict the community seems to be about recording without consent. Some people go to burns to be able to completely disconnect from their usual lives without fear that there will be any reprisal for legal/maybe-illegal-but-harmless activities they might do there, and the potential of being recorded can put a serious damper on that feeling of freedom.
I don't see evidence of facial recognition.
Face recognition means computing which individual from some other database of people a particular face belongs to.
There’s also face tracking — detecting a face in an image and then tracking the same face across subsequent images. Which is often implemented by using a face recognition approach, but without any predefined catalog of people — you just dynamically fill up your face database as faces appear in the image sequence / video source.
Recognition implies associating the faces with an ID.
It detects objects and gives you the bounding box. Then you draw a square on it and add a label.
No fancy LLM needed, just old fashioned machine learning models.
God damn those are 12 great songs!
I also hadn’t really clued in to just how political they were until seeing their visuals, which I also thought added a lot. Surely not everyone’s cup of tea though.
Bear in mind Beth made "Out of Season" apart from Portishead several years before the release of "Third." I wouldn't think her recent solo work indicates a split.
But yes. They do need new material dammit.
I attended one in Paris this year. Not only it was exactly the same show as few years ago including the decorations and the scrolling text wall but the sound was horrible. I couldn't hear anything.
I would be better off if I just stayed home and listened to the recording.
The best concerts are breakthrough concerts for new bands (first or second album), and then the greatest-hits type concerts that are 10 years after the last album. Every concert that is a tour with album 3-4-5 is usually pretty meh.
Have they done this again with an updated system?
And that's part of the performance. You don't get to choose what companies do with your personal data.
It’s hard to explain the concept of surveillance and its effects to laypeople. And the corporations absolutely know that.
And it isn't identifying the people or anything. It's putting some meaningless adjective like "Resourceful" below them.
Have seen this headline a few times and thought it was actually novel and demonstrative of some face database or something, but instead it's just a surveillance gimmick. Put a bunch of generative AI face loops with bounding boxes and adjectives.
"AI assists in refining our editorial process, ensuring that every article is engaging, clear and succinct."
One thing I hope we'll see in the future on these types of articles is the ability to view the original prompt. If your goal is to be succinct, you can't get much more succinct than that.
What is sadly rather ironic is the author's first name, "Al" looks like AI when stylised in the article's font.
Writing for the last 14 years and for GadgetReview since 2017; Managing Editor since 2018.
I think this assumes a very limited scope of how AI gets used for these. As if the article is a one and done output from a single prompt. I can imagine many iterative prompts combined with some copying and pasting to get an hour’s worth of copy in five minutes.
Would it matter if the same prompt gives different output? You couldn't verify it.
How likely are you to just trust, let alone know for sure, whether or not the text I showed you is actually what I fed to the llm?
a smart person can make ChatGPT sounds completely authentic, and a very boring and middle of the road writer who uses em-dashes can make themselves sound completely inauthentic. it's not like LLMs got their style from nowhere
as far as I'm concerned, as long as the factual information has been curated by a human, I don't give a shit
How could you possibly tell? I've been playing around with AI detectors, putting in known all-human samples, known all-AI samples, and mixed samples.
The only thing it's gotten right is not marking a human sample as 100% AI (but it marked one of the AI samples as 100% human).
Having such a mark would be a witch-hunt for sure.
Nothing personal, but you do seem to have a nice education. US ?
In fact, I recall many songs from The Matrix being played nonstop back in my teenage gamer IRC days. Maybe even by others than just me
The headline is perfectly parseable, unless most of the headlines on HN or BBC. The fact that it says "Band concert" shall be selfexplanatory.