We’ve reached a stage, where it would be advisable to not release recent photos of yourself, nor any video with sound clips to public, unless you want an AI fake instaperson of yourself starting to reach out to member of your externally visible social network, asking for money, emergency help, etc.
I guess we need to have an AI secretary to take in all phonecalls from now on (spam folder will become a lot more interesting with celebrity phone calls, your dead relative phoning you etc)
Hopefully, we will soon enter the stage where nobody believes anything they see anymore. Then, you no longer have to be afraid of being misinterpreted, because nobody is listening anymore anyway. Great time to be alive!
Luckily there’s a “solution” to that: Just don’t use the internet for dialogue anymore.
As someone that grew up with late-90’s internet culture and has seen all the pros and cons
and changes over the decades, I find myself using the internet less and less for dialogue with people. And I’m spending more time in nature and saying hi to strangers in reality.
I’m still worries about the impact this will have on a lot of people’s ability to reason however. “Just” Tik Tok and apps like it has already had devastating results on certain demographics.
That bit "... there's a "solution"" - does it keep working in societies where there are mega corps pushing billions into developing engaging, compelling and interesting AI companions?
That's why I put it in quotation marks because it is a solution that will remain available, simply because the planet is really big and there'll always be places on the fringes. But it doesn't really solve the problem for society at large, it only solves it for an individual. But sometimes individuals showing other ways of living helps the rest of society see that there's choices where they previously thought there were none.
I don't know why anyone thinks this will happen. You can obviously write anything you want (we have an entire realm of works in this area that everyone knows about, fiction) and yet huge amounts of people believe passed around stories either from bad or faked media sources or entirely unsourced.
I'm not saying either you or the parent commenter is right or wrong, but fiction in books and movies are clearly fiction and we consume it as such. You are right that some people have been making up fake stories and others (the more naive) have been quick to believe in those false stories. The difference now is that it's not just text invented and written by a human, which takes time and dedication. Now it's done in a second. On top of that it's easy to enhance the text with realistic photos, audio and video. It becomes much more convincing. And this material is created in a few seconds or minutes.
It's hard to know what to believe if you get a phone call with the voice of your child or colleague, and your "child"/"colleague" replies within milliseconds in a convincing way.
I agree it's fundamentally different in application which I think will have a large impact (just like targeted advertising with optimisation vs billboards), but my point is that given people know you can just write anything and yet misinformation is abound - I don't see how knowing that you can fake any picture or video or sound leading to a situation where everyone just stops believing them.
I think unfortunately it will massively lower the trust of actual real videos and images, because someone can dismiss them with little thought.
Be glib, but that is one way for society to bring privacy back-and with it shared respect. I think of it as the “oh everyone has an anus” moment. We all know everyone has one and it doesn’t need to be dragged out in polite company.
I'm not sure if people work like that — many of us have, as far as I can tell for millennia and despite sometimes quite severe punishments for doing so, been massive gossips.
What you see will be custom tailored to what you believe, and your loyalty will be won. Do what the AI says and your life will be better. It already knows you better than you know yourself. Maybe you're one of those holdouts who put off a smartphone until life became untenable wihout it. Life will be even more untenable without your AI personal assistant/friend/broker/coach/therapist/teacher/girlfriend to navigate your life for you.
I think for most people it's far too late, as there exists at least something on the internet and that something is sufficient - photos can be aged virtually and a single photo is enough, voice doesn't change much and you need only a tiny sample, etc.
And that's the case even if you've never ever posted anything on your social media - it could be family&friends, or employer, or if you're ever been in a public-facing job position that has ever done any community outreach, or ever done a public performance with your music or another hobby, or if you've ever walked past a news crew asking questions to bystanders of some event, or if you've ever participated in some contests or competitions or sports leagues, etc, all of that is generally findable in various archives.
> photos can be aged virtually and a single photo is enough
I'm sure AI-based ageing can do a good enough job to convince many people that a fake image of someone they haven't seen for years is an older version of the person they remember; but how often would it succeed in ageing an old photo in such a way that it looks like a person I have seen recently and therefore have knowledge rather than guesses about exactly what the years have changed about them?
(Not a rhetorical question to disagree with you, I genuinely have no idea if ageing is predictable enough for a high % result or if it would only fool people with poor visual memory and/or who haven't seen the person in over a decade.)
I feel like even ignoring the big unknowns (at what age, if any, will a person start going bald, or choose to grow a beard or to die their hair, or get a scar on their face, etc.) there must be a lot of more subtle but still important aspects from skin tone to makeup style to hair to...
I've looked up photos of some school classmates that I haven't seen since we were teens (a couple of decades ago), and while nearly all of them I think "ah yes I can still recognise them", I don't feel I would have accurately guessed how they would look now from my memories of how they used to look. Even looking at old photos of family members I see regularly still to this day, even for example comparing old photos of me and old photos of my siblings, it's surprising how hard it would be for a human to predict the exact course of ageing - and my instinct is that this is more down to randomness that can't be predicted than down to precise logic that an AI could learn to predict rather than guess at. But I could be wrong.
Maybe it's Europeans posting this kind of stuff where they have much stronger privacy laws, but if you're in the US this is all wishful thinking.
Do you shop in large corporate stores and use credit cards? Do you go out in public in transportation registered to you?
If yes, then images and habits of yours are being stored in databases and sold to data brokers.
And you're not even including every single one of your family members that use internet connected devices/apps that are sucking up all the data they can.
I was just asking about the ability of photo aging software, not commenting about privacy at all. Though yes, I am thankfully in Europe (but there are recent photos of me online).
But don't disagree with you - in a different comment that was about privacy, I (despite living under GDPR) suggested that for offline verification with known people it's better to choose secrets that definitely haven't been shared online/anywhere rather than just choosing random true facts and assuming they couldn't have been found out by hackers: https://news.ycombinator.com/item?id=40353820
> I guess we need to have an AI secretary to take in all phonecalls
Why not an AI assistant in the browser to fend all the adversarial manipulation and spam AIs on the web? Going online without your AI assistant would be like venturing without a mask during COVID
I foresee a cat-and-mouse game, AIs for manipulation vs AIs for protection one upping each other. It will be like immune system vs viruses.
I'm paranoid enough that I now modulate my voice and speak differently when answering an unknown phone call just in case they are recording and building a model to call back a loved one later. If they do get a call, they will be like, "why are you talking like that?"
But why not just make up a secret word to use with your beloved ones in critical situations. In case of ..., one needs to know that secret. Otherwise, FAKE! Gotcha!
The problem here is you're assuming your family members aren't idiots, this is your first mistake.
Chances are they've already shoved some app on their phone that's voice to txting everything they say and sending off somewhere (well lower chance if they have an iphone).
Modern life is data/information security and humans are insanely bad at it.
By chance, they are noobs but not idiots, because they ask me on everything - they don't need Google, I know everything hahah
I don't think it's a problem to find a word or a sentence or a story - whatever - that's commonly used by everyone on daily basis but in different context. That's not a problem by itself :) try it
For the idiots, it is still possible to find a word. They may be idiots, but still, they work and live on their own. They coming along in life. So, it's up to the smarter one to find a no-brainer solution.
I am confident and believe nothing and no one is stupid enough not to be able to adapt to something. Even if it's me, who'll need to adapt to members with less brain.
This is my biggest gripe against the telecom industry. Calls pretending to be from someone else.
For every single call, someone somewhere must know at least the next link in the chain to connect a call. Keep following the chain until you find someone who either through malice or by looking the other way allows someone to spoof someone else's number AND remove their ability to send the current link in the chain (or anyone) messages. (Ideally also send them to prison if they are in the same country.) It shouldn't be that hard, right?
Companies have complex telecoms but generally want the outside as one company number. Solution, the sender send a packet with the number they should get perceived as. Everyone sends this on. Everyone "looks the other way" by design haha
So what, gate that feature behind a check that you can only set an outgoing caller ID belonging to a number range that you own.
The technology to build trustable caller ID has existed for a long time, the problem is no one wants to be the one forcing telcos all over the world to upgrade their sometimes many decades old systems.
I guess we need to have an AI secretary to take in all phonecalls from now on (spam folder will become a lot more interesting with celebrity phone calls, your dead relative phoning you etc)