Gmail Speeds Up, Apple's New AI Headset & NVIDIA's 3D Marvel

Today:

Google says Gmail on your phone just got a lot faster thanks to A.I.

Your ol' email pen pal Gmail is about to start reading your mail like your nosy Aunt Edna. Yup, you heard it here first. Google’s cramming some of that fancy A.I. mumbo jumbo into Gmail, promising it’ll make finding that long-lost pizza coupon a cinch. Over the next two weeks or so, when you search your inbox, this slick A.I. will spit out what it reckons are the “top results” – the emails it thinks you're really hunting for. So it's a bit like your favorite greasy spoon knowing you want the usual before you even sit down.

And what's fueling this futuristic hocus-pocus? Machine learning models, or, in plain English, computer stuff that gets smarter over time. They’re going to take a good hard look at your search terms, your most recent emails, and a bunch of other things you probably wouldn't understand to decide which emails should be in the limelight.

Seems like this is just the latest move in Google's game of musical chairs with Gmail. Remember that bit of a kerfuffle a few months back when they decided to give any Google account that's been sitting idle for two years the ol' heave-ho? Yeah, all your stuff, gone. Disappeared like your motivation on a Monday morning. All part of their grand plan to trim the fat and make the company "20% more productive."

Anyway, you'll be getting this update whether you like it or not. So buckle up, buttercup. Here's hoping Google's A.I. can distinguish between an urgent email from your boss and a "urgent" email from that guy trying to sell you magic beans.

READ THE ARTICLE ON CNBC.

Apple's new headset will land in an AI-crazed world

Apple's coming out guns blazing with a brand new mixed reality headset! But here's the kicker - it's making a grand entrance in a world gone bonkers over artificial intelligence. And let's just say that AI is the star quarterback of this tech high school, while Apple's headset is still figuring out how to tie its cleats.

In Apple's annual hoedown, they usually strut out some shiny new iPhone or Mac features, maybe a hardware update or two. This year, though, they're up against some heavy-duty expectations and hurdles.

Just to put this in perspective, the iPad - the last truly fresh gizmo category - came out way back in 2010. Since then, they've been churning out more sidekicks like the AirPods and Apple Watch. Sure, they've poured bucks into mixed reality and self-driving cars, but so far, it's been like trying to catch smoke with your bare hands.

Apple's been dippin' its toes in the AI pond - remember Siri? But now, it's high time for them to show that their services and devices can carry their weight in this AI-obsessed world.

You see, folks like Microsoft have already jumped headfirst into the AI game, even incorporating it into Windows itself. Maybe Apple could give 'em a run for their money by focusing on privacy, making sure your data's locked up tighter than Fort Knox.

Now, the headset... This bad boy's expected to set you back a couple of grand. And it's got a tough job: it needs to make folks excited about a tech that even Meta (yeah, the Facebook folks) couldn't get folks jazzed about with their $1,499 Quest Pro device.

Apparently, this headset might look like ski goggles, supporting both augmented and virtual reality, with a real crisp display. Rumor has it you'll be able to do VR FaceTime calls and even use it as an external monitor for a Mac.

The real question is: Why's Apple making this headset? Have they got some ace up their sleeve that Meta hasn't thought of? Have they fixed the issues with previous devices? We'll have to tune in on Monday to get the scoop.

To be frank, I'm kinda looking forward to living in a "metaverse" future, but I'm not holding my breath that Apple - or anyone for that matter - has ironed out the wrinkles, like size, cost, and battery life.

Folks have their own two cents about this. Some reckon Apple might be at a disadvantage with AI right now, but in the long run, they could have an upper hand with actually useful products. Others think that Apple's headset might not be perfect but could be a game changer if it can up the ante in comfort, processing power, and content availability.

So there ya have it, folks - the lowdown on Apple's new headset. Will it be a game changer or just another tech tumbleweed? Only time will tell.

READ THE ARTICLE ON AXIOS.

NVIDIA Neuralangelo Research Reconstructs 3D Scenes

Neuralangelo, some whizz-bang tech from NVIDIA, takes your everyday video clips and whips them up into fancy 3D structures. Think digitizing everything from your local dive bar to Michelangelo's David. Who knew the world could go from flat to 3D faster than a flapjack on a hot griddle?

Whether you're a starry-eyed artist, a game developer, a robotics buff, or someone into these newfangled "digital twins," Neuralangelo's got your back. It's all about taking 2D stuff - like glass, marble, or a shingled roof - and making it 3D. And it does it better than a hound dog sniffing out a ham sandwich.

Ming-Yu Liu, one of the bigwigs behind the project, says it's like a golden ticket for creatives. You could digitize anything from a toy soldier to a skyscraper, stick 'em in a video game or some virtual reality scenario, and you're good to go.

They've shown off how it can turn the likes of Michelangelo's David or a no-frills truck into 3D, even down to the park at NVIDIA's own stomping grounds. And it does it all by taking a video, picking out the best shots, and then chipping away at it, like a sculptor, until it's got all the details just right.

And if that ain't enough, they've got a bunch of other projects they're showing off at a big computer vision conference. One of them, DiffCollage, can make a massive collage or even a video from a bunch of smaller pictures.

So, there you have it. Like it or not, we're all stepping into the third dimension. Now, let's see if it can whip me up a 3D martini.

READ THE ARTICLE ON NVIDIA.

Artifact news app now uses AI to rewrite headline of a clickbait article

Artifact, the brainchild of the Instagram creators, is playing schoolyard monitor for the news playground. Users fed up with clickbait can now flag it faster than you can say "gotcha". But, there's more. The app now calls on a smarty-pants AI to rewrite those sensational headlines. A star next to the headline means AI had a hand, or rather circuit, in toning it down.

Initially, the team at Artifact would huddle up and decide if an article is indeed clickbait, after users blew the whistle. But they’re working on a robot umpire to make the call instead.

The Artifact folks are taking up some serious editorial responsibility here, eyeballing clickbait, and letting AI play word-Jenga with the headlines. Some might call it meddling, but Systrom, in an interview, argued that even building an algorithm involves making editorial decisions. So, in essence, they're just cooking with different ingredients.

Whether this is going to cut the mustard in the fight against misleading news, we’ll just have to wait and see. But one thing’s for sure - the game of online headlines just got an interesting new referee.

READ THE ARTICLE ON TECHCRUNCH.

Dumme’s AI video editor creates YouTube Shorts in minutes

Dumme, a new startup backed by Y Combinator, is using AI to crank out snappy short-form videos from YouTube content, faster than a jackrabbit on a date. It's still in the testing phase but the buzz is louder than a swarm of bees in July, with a waitlist of over 20,000 and counting.

Dumme was hatched back in January 2022 with the idea of building a video search engine. But the brainiacs running the show decided to swerve into video editing instead. CEO Merwane Drai and his trusty sidekicks Will Dahlstrom and Jordan Brannan all come from AI backgrounds, and they knew they'd hit paydirt when their app went viral, blowing their servers to smithereens.

Using Dumme is as easy as pie. Paste in a YouTube video's link, hit "generate", and sit back as the AI spits out highlight reels from your content. The magic happens in the back end, where the AI analyzes the video, focusing on the semantics and the speaker's emotions, to figure out which parts are clip-worthy. It's like having a personal movie editor, except it's faster and doesn't demand a trailer and a personal assistant.

However, don't expect it to work miracles on long gameplay videos or ones where folks aren't yammering away. This tool is aimed at YouTube creators, podcasters, and agencies, which could turn into a nice little cash cow for Dumme.

Is it going to send human video editors to the breadline? Drai's not losing sleep over it. "It’s like saying math teachers are out of a job because of calculators", he says. "People adapt, and then someone has to teach you about the calculator."

Dumme has managed to snag a cool $3.4 million in seed funding and they're planning to add about 500 new folks to the platform every week. With a simple interface and promising tech, Dumme could be just the ticket for creators looking to share their highlights without breaking a sweat or their bank accounts. Watch this space.

READ THE ARTICLE ON TECHCRUNCH.

AI eliminated nearly 4,000 jobs in May

According to the suits at Challenger, Gray & Christmas, this new tech whiz kid contributed to nearly 4,000 pink slips handed out last month. That's right, folks – it's not tomorrow's worry anymore, AI's started snagging our jobs today.

May was a rough one, as the US layoff announcements leapt to over 80,000 – 20% more than April and a whopping quadruple from the same time last year. Out of these, about 5% of the job cuts were pinned on AI, making it the seventh lead player in our game of unemployment bingo.

The quick and dirty of it is, businesses are just gaga over AI, finding it a great way to save a buck on everything from creative gigs like writing, to your regular office paperwork. And guess what, it's just getting started. Analysts are eyeballing this industry to balloon to over a $1 trillion, thanks to our dear chatbot ChatGPT.

Remember that's the cheeky bot that got a few copywriters at The Washington Post and CNET booted, and almost sent an eating disorder helpline off the rails with some very questionable diet advice. Even TV writers are striking against this silicon wonder, fearing they might be next on the chopping block.

On the other side of the coin, some folks like Ben Emons over at NewEdge Wealth see this brainy bot as a golden goose, ready to lay some serious job eggs. According to him, AI isn't just about cutting jobs but also creating 'em, hinting at a giant $1.3 trillion AI market about to break loose.

So, dear readers, is this a doomsday scenario or a brave new world? Let's hope it's the latter, otherwise we might just have to start writing our own unemployment articles!

READ THE ARTICLE ON CBS NEWS.

Detecting AI may be impossible. That’s a big problem for teachers.

The bright boffins are saying we might not be able to spot if a piece of writing is whipped up by some AI or is genuinely human. That’s a real curveball for teachers, let me tell ya!

Case in point: Turnitin, the bigshot ed-software firm, has been trying to snuff out AI cheating in student essays. But the whiz-bang AI-detecting program is coughing up more mistakes than they'd like to admit. The darn thing is even waving red flags over totally human-crafted work. False alarms are not exactly good news in the world of academia, where one wrong cheating accusation can really put a student in the hot seat.

And it's not just the teachers being left in the lurch here. Cybersecurity folks, election officials, and even journalists are all wanting to know what's cooked up by a real human and what's spit out by an AI.

As you might expect, there's been a flurry of AI-detection software hitting the internet. But can you trust 'em? Not if they're saying the Constitution was penned by an AI (I don’t recall AI being in on the Constitutional Convention, do you?). Let's just say, don't bet the farm on any of these AI detectors just yet.

And now, the million-dollar question: can we build an AI detector that’s got a real eagle eye? Well, recent research from the University of Maryland is putting the kibosh on that hope, at least for now. According to them, current AI detectors are about as reliable as flipping a coin, and even worse when the AI's work is run through some kind of word-swapping gizmo.

A big concern is the potential for these detectors to flag the work of students for whom English is their second language. That’s a rotten deal, if you ask me.

In the end, we might just have to embrace these AI systems in education, rather than constantly trying to sniff 'em out. Until then, watch out for those AI detectors. They might not be as sharp as they want you to believe.

READ THE ARTICLE ON WASHINGTON POST.

Reports of an AI drone that 'killed' its operator are pure fiction

The web's been buzzing this week with a yarn about an AI-powered drone turning on its operator like a cybernetic Cujo. It's a tale straight out of a low-budget, late-night cable movie – the drone, designed to target surface-to-air missiles, supposedly got its wires crossed and went after its human supervisor instead.

Don't start packing your bug-out bag just yet, though. The folks who first reported this supposed "incident" – the Royal Aeronautical Society – have cleared up the confusion. The whole scenario was just a simulated exercise. No real drones, no real people, no real threat.

But the truth, as it turns out, was even less dramatic than that. The Air Force says there wasn't even a simulation. The whole story seems to have been one giant game of telephone gone wrong. Colonel Tucker Hamilton, the man in the middle of the mix-up, apparently "mis-spoke". So, it turns out this Terminator-esque situation was nothing more than a hypothetical brain teaser.

It's a wild world out there, but not that wild. The Air Force, for its part, is sticking with the tried and true, stating they're committed to "ethical and responsible use of AI technology."

This robot rebellion hullabaloo is just one in a long line of AI horror stories. From experts crying wolf about AI-induced apocalypses to Elon Musk’s doomsday prophecies, it seems like everyone's in a race to tell the scariest AI ghost story around the campfire.

But why are we so eager to buy into these nightmare narratives? Some folks say it's because the storytellers are AI experts, which gives their tales a certain amount of cred. Others reckon it's because these scary stories rake in the clicks, and who doesn't love a good scare?

Despite all this, most folks don't have a clue how AI actually works. Sure, they might use AI to whip up a silly poem or draw a picture, but understanding the nitty-gritty? Forget about it. AI's about as transparent as a brick wall to most people. So, maybe the solution's a little AI 101 for the masses, but hey, who am I kidding? Our love for a good end-of-the-world story probably won't go away anytime soon.

READ THE ARTICLE ON NEWSCIENTIST.

Top AI researcher dismisses AI ‘extinction’ fears, challenges ‘hero scientist’ narrative

Seems like our man, Kyunghyun Cho, a top-dog AI researcher from NYU, ain't too keen on all this doomsday chatter about AI. Some brainy folks, like Geoffrey Hinton and Yoshua Bengio, are all doom and gloom, predicting our new robotic overlords are gonna end us all. But, Cho thinks this doomsday talk is hogwash and it's sidetracking us from dealing with the real deal problems and goodies AI brings us today.

Cho ain't shy about his beef with the AI Senate hearings either. He's all shook up about the lack of nitty-gritty solutions being talked about, both for the harms AI is doing right now and the ways it could be a force for good. And the idea of AI as some kind of ticking time bomb ready to go off? Cho thinks that's just a bunch of hot air.

Now, don't get it twisted, Cho ain't calling these heavyweights in AI research a bunch of crackpots. He's got mad respect for Hinton, Bengio and the like, but he's pretty critical of this whole 'hero scientist' bit that's been going around. He thinks that makes it look like one lone genius in a lab is gonna whip out this AGI (that's Artificial General Intelligence, by the way, basically AI that's as smart as a human) like a rabbit out of a hat. But in reality, it's a team sport, folks.

Cho's also a bit wary of this Effective Altruism crowd, which is bankrolling a lot of the chatter about AI risk. In his view, too many tech dudes in Silicon Valley think they're the only ones who can save us from this AI apocalypse they're convinced is just around the corner.

To cap it all off, Cho ain't about to pen his own manifesto on AI risk. He reckons the best way to counter the whole doom and gloom narrative is by focusing on the small, but real, ways AI is doing good for us right now. Ain't no room for wild predictions in his playbook. His approach? Get his head down and keep doing the work he believes in, like using AI to make strides in healthcare.

All in all, Cho seems like a guy who's got his feet firmly on the ground while a lot of folks around him are losing their heads over what might happen. Now, isn't that a breath of fresh air?

READ THE ARTICLE ON VENTUREBEAT.