Artificial Intelligence. Yes or No?

Artificial Intelligence (AI) is certainly making more inroads into everyday life, to the point where we don’t necessarily have to push a button any more, because all we have to do is speak to the compact little unit that makes it all happen and it all happens! Interesting.

So we no longer have to get off the couch if we don’t want to, just speak a command and the TV will come on (could this be the end of losing the remote?), music will play, the phone will call whoever we wish to speak to and the home AI system will even lock doors, turn on lights and send footage from our home security cameras to our phones so we can check what goes on when we’re not there.

It almost rules out having to think too much, doesn’t it? Let alone having to physically do something, and while this may make those who use it (and more people are now) the envy of those who want to but can’t afford the setup yet, I sometimes wonder just how good an idea it really is. Do I really want a little gadget running my life for me? Sure, they look cute and everything, but…is this what I really want?

Okay, so it’s fun and it’s really convenient. Via an AI home system you can send the little vacuuming robot on a whiz around the house while you check the recipe the system has brought up on your kitchen device, or set the kettle to boil while you spend some downtime with a downloaded novel. It can’t bring your coffee to you yet, but give it a couple of years and there will be something that will. It will probably make it for you first!

Hang on, aren’t there coffee machines that already make it for you, from go to whoa, now? Whatever, they still can’t bring it to you at this point. Anyway…

So then, your house is now the coolest one on your block because it does everything for you and all you have to do after a hard day’s slog is come home, walk in and drop into the massage chair. Well sure, that sounds all very nice, but what will you do in regards to general physical activity? You know, as in when you used to move about from A to B to C around the house doing actual things for yourself? Oh that’s right, your AI gadget probably talks to your exercise machines (it definitely will if you have a Google Nest Audio) and all you need to do is verbally direct it and it will set them in motion for you. Heaven forbid you should have to flick a switch or press a button.

Okay, fine! That’s why you went all AI in the first place so that you don’t have to do any flicking or pressing! Whatever!

But the big question for me is this…what happens should your system develop a glitch? Or gets hacked? Like, it cuts off your phone, sets the exercise bike to manic and tells your car not the start? Then it prevents you from initiating the manual override, locks you all inside or worse, locks you all out of your house altogether? You are probably thinking that is just ridiculous but if it’s smart enough to run everything in your house, run your phone and your car, it just might be smart enough to play silly buggers with you, or worse, your AI system decides it doesn’t like you anymore and goes all Twilight Zone on you.

Or the hacker does. Either way, it’s not good.

Personally, I like perusing the old recipe books, the physical ones, when I’m in the mood to create magic in the kitchen.  I don’t have issues with pressing buttons or flicking switches, with walking over to the bookshelves and physically selecting a real book to read. I am not put out by having to turn a key in a lock or trot around with the vacuum cleaner and feel comfortable with not wanting or needing to speak to a gadget to make things happen.

Perhaps I’m just too smart for AI.

 

 

 

Beware What You Share

Despite the popular trend to share selfies and all sorts of personal information on social media, it is something I have never done, because in the back of my mind was a little niggle that made me pause, and then opt Not to plaster my Facebook page with pics of myself, posts about where I am at any particular time, or any personal information or issues on what I am doing or experiencing. If anything, I have kept it really light.

And now I am soooo glad I did that. Why? Because of Hoan Ton-That (born Cam-Hoan Ton-That) and his company, Clearview AI.

A very clever boy, Hoan Ton-That grew up in Lyneham, Canberra, where he attended Lyneham High School, finishing Year 10 in 2003. He completed Years 11 and 12 at Hawker College and from there went on to the Australian National University (ANU) to study for his Bachelor of Information Technology degree. Ton-That registered a company in 2006 named Software for Mom Pty Ltd and appeared to be on his way to a bright future. Then he dropped out of university in 2007, before finishing his studies, and took himself off to Silicon Valley.

The first hiccup in what appeared to be the start of a promising career was the accusation that surfaced in America in 2009, which alleged Hoan Ton-That was behind two phishing scams, and in a court case currently running in the US against CLearview AI, it has been alleged that Ton-Than was also involved in other companies, including ViddyHo.com (another alleged phishing site which conned users into sharing access to their gmail accounts, which allowed the site to then generate spam to those on the users’ contact list). Not good.

The other alleged phishing site apparently connected to Ton-That was Fastforwarded.com, which attempted to extract passwords from users by fraudulent means. Also not good.

And then there’s Clearview AI. This company was founded in 2017, initially under the name Smartcheckr, and is a powerful facial recognition technology which sells access to the site to various law enforcement agencies, commercial businesses and a number of government agencies who can then upload images and, with the aid of artificial intelligence, make comparisons to the huge store of photographs in Ton-That’s database. If they find any matches, the database will then provide all the information it has available on those people; information that has been mainly accessed from their social media accounts.

But according to Clearview AI’s lawyer, Tor Ekeland, it is all above board and Clearview AI’s primary purpose is “focused on providing it’s groundbreaking and effective technology to assist law enforcement in the US in identifying perpetrators of horrific crimes such as trafficking and child abuse…”. He goes on to say more of the same but disputing that, I suppose, would be the class action Clearview AI is also facing in Illinois for breaching privacy provisions, a similar action in Virginia also and as a result, the police force in New Jersey has been advised to stop using it.

Here in Australia, it has surfaced that police officers in Victoria, South Australia and Queensland have also held a number of registered accounts with Clearview AI. The Federal Police have too. So how did they get onto it? Clearview AI did offer a free pilot trial and the belief is that most of the officers with access had signed up for the freebie, as opposed to signing up for an account, which comes with a very expensive yearly fee.

An investigation into Clearview AI has been launched by The Australian Information Commissioner and the site has been disabled in Australia. That’s good news, but what isn’t all that great is that Hoan Ton-That has already built a database of over three billion photographs, almost all of which have been “scraped” from social media accounts via computer programs more commonly known as spiders, or web crawlers. Many Australians were scraped. Twitter, Google and Facebook have apparently served the company with “cease and desist” communications and Apple has also blocked the app from its store.

But that’s a bit like locking the stable door after the horse has bolted, isn’t it? Ton-That already has that enormous database and most of it was obtained with neither the knowledge, nor the permission of those he scraped, and he is happy to sell that information to the highest bidder. The photos and accompanying information were scraped intentionally and unfortunately, anyone who is good for the fee could get access to it, not just legitimate law enforcement agencies. Private commercial facial recognition databases are not bound by the same laws that apply to similar programs under government control and that leaves the door wide open to undesirables getting a hold of it, and once they have it… Fortunately, I am highly unlikely to be an attractive candidate for this site because of the lack of pics and info, but unfortunately I can’t say the same with any real assurance about some of my online friends and family members and that’s a concern.

So just be careful what you post, because there is probably another Hoan Ton-That out there somewhere. Possibly more than one.

And they’re looking for someone just like you.