An urgent wake-up call from leading figures in AI research and development. They warn of an ‘extinction’ risk related to AI advancements, advocating for preventative measures and thoughtful regulations.


AI industry and researchers sign statement warning of ‘extinction’ risk

A bunch of big-brained folks, tech honchos, and even some celebs got together and said, “Hang on, this AI stuff could really blow up in our faces.” They scribbled a note saying the risk of us all getting snuffed out by AI should be right up there on the worry-list, shoulder to shoulder with nasty bugs and nuclear kabooms.

Among the scribblers, we got Sam Altman, the head honcho at OpenAI; Geoffrey Hinton, the granddaddy of AI; and a choir of top dogs from Google DeepMind, Anthropic, and Microsoft. And for some reason, the climate champion Bill McKibben and the singer Grimes also hopped on the bandwagon.

Now, don’t get your knickers in a twist just yet. These tech wizards say we’re still a country mile away from the sort of self-thinking AI you see in those sci-fi flicks. Today’s top-drawer chatterbots just spit back out the info they’ve been fed; they’re not going rogue on us…yet.

But with all the fuss and dollars being thrown at AI these days, folks are hollering for some ground rules before things get out of hand.

This all comes on the heels of the success of OpenAI’s ChatGPT, which has the tech world trying to outdo each other in the AI department. Meanwhile, lawmakers and other worrywarts are waving red flags about how these new-fangled AI chatbots could spread baloney and snatch up jobs.

Geoffrey Hinton, who’s done a bunch of groundwork for AI, ditched his gig at Google to sound the alarm on the tech, saying these AI gizmos are getting too big for their britches.

Despite this doom and gloom, Dan Hendrycks from the Center for AI Safety tweeted, “Hey, we can handle more than one problem at a time.” He’s saying we can’t just focus on what’s biting us in the butt right now, we also need to keep an eye on potential future hiccups. ‘Cause, you know, not doing that would be just plain dumb.


Google DeepMind introduces Barkour, a benchmark for quadrupedal robots

Google DeepMind’s cooked up something called “Barkour.” A kinda playground slash report card for those four-legged robots – you know, like that fancy Boston Dynamics’ Spot that everyone’s jawing about.

Quadrupeds – fancy word for four-leggers – have been strutting their stuff in labs, industries, even on soccer fields. Some, worryingly, are playing RoboCop too. As the two-legged robot wannabes are still figuring out their left foot from their right, these four-leggers are out there making hay.

DeepMind, Google’s brainy kid, fresh from adopting the flagging Everyday Robots team, has whipped out a new research paper. They’re pitching “Barkour” as a kinda SAT test for these metal mutts, seeing how well they can navigate obstacles and such.

The whole shtick seems to be inspired by man’s best friend. They set up an obstacle course, plonked a hot dog (the dachshund type, not the ballpark frank) in it, and watched how it did. The robo-dogs had to do the same – hop, skip, and jump over the hurdles in about 10 seconds, same as Fido. A no-nonsense 0 to 1 scoring system – either you make it, or you don’t. Slacking off or playing hooky with the obstacles racks up penalties.

Google’s crowing about how Barkour’s a real game-changer for sussing out the agility of these robo-dogs. Apparently, the mechanical mutt they tested managed to pick itself up after a face-plant and hoof it back to the start. So, there you have it. “Barkour” – Google’s new report card for robo-pooches. Just don’t expect them to fetch the paper anytime soon.


MyHeritage debuts Reimagine, an AI app for scanning, fixing and even animating old photos

So, MyHeritage just released an app called Reimagine to help folks keep their family photos in check. We’re not just talking a quick scan-and-save here; we’re talking touch-ups, fix-ups, and even bringing those faces to life. Yep, the app can even animate the faces, just like that Deep Nostalgia trick they pulled off before.

Here’s how it works. You snap a pic of your old photo album, and the app will automatically crop out the individual pics for you. Plus, you can add names, dates, and places for easy finding later. Got a shoebox full of snapshots? They got you covered.

But, let’s be honest, Google did the scanning thing five years back. The cool beans about Reimagine, however, is the fix-up and spruce-up of your pics. It can make colors pop, patch up scratches, and even give low-res images a facelift. And the icing on the cake – you can animate old photos and add a voiceover. Ghostly? Maybe. Cool? Definitely.

Test-driving the app, we found it mostly lived up to the hype. The only hiccup was the AI failing to fix a glare issue in one pic. So while the touch-up results might not blow your socks off, they’re a definite step up from a blurry old mess.

Reimagine’s color restoration also needs a second look before we give a thumbs up. But, hey, MyHeritage has been using this tech since early 2020, so they have had time to tweak and twiddle. Just note that it’s not about turning black-and-white into color, but about rejuvenating faded colors in old color photos.

As for the dollars and cents, the app is free to download, but if you want the whole shebang, you’ll need to shell out $7.99 monthly or $49.99 yearly. Not too shabby for bringing those dusty memories back to life. And if you’re wondering, they’re putting a watermark on manipulated pics to keep things above board. So, check it out and see if Reimagine can breathe some life back into your old family snapshots.


Introducing Charlotte AI, CrowdStrike’s Generative AI Security Analyst: Ushering in the Future of AI-Powered Cybersecurity

Time to chew the fat about CrowdStrike, y’know, those cybersecurity whizzes. They’ve been playing footsie with artificial intelligence (AI) for over a decade now. Got them a new shiny toy they’re callin’ Charlotte AI, a real humdinger of a security analyst, built right into their Falcon platform.

What’s that mean for you? Glad you asked. Charlotte’s basically your personal safety guide in the wild west of cyberspace. Say you’re a big-shot CEO or a tech newbie, you just shoot a question at Charlotte in plain ol’ English like, “What’s the big risk to our computers?” and wham! She’ll spill the beans, easy as pie.

Three big ways this gal Charlotte’s gonna help you out:

  1. She’s a real peach for helping everybody become a cyber whizz. Need to impress the big wigs at the board meeting? Just ask her some questions and she’ll sort you out with what you need to know about your company’s cyber risks.
  1. She’s also aces at helping greenhorn IT folks. She can guide them through their security duties like a seasoned vet, answering questions about vulnerabilities and threats lickety-split.
  1. For the big guns in security, Charlotte’s like an extra pair of hands. She can do the grunt work like data collection and basic threat search so you can focus on the fancy stuff.

Now, you might be wondering, how’s she do all this? Well, it’s all thanks to CrowdStrike’s monster truck of data. They’ve got info on every dirty trick cyber crooks use, data from their Falcon platform, and the expertise of their top-notch team. Put that all together and you’ve got Charlotte – a real cyber whiz in your pocket.

So, that’s the skinny. CrowdStrike’s cooked up a real game-changer with Charlotte AI. She’s set to make life a whole lot easier for everyone dealing with cybersecurity, no matter if they’re new to the rodeo or an old hand. And ain’t that a breath of fresh air?


Nvidia is now a $1 trillion company thanks to the AI boom

Alright, so Nvidia’s just been inducted into the ‘Trillion Dollar Club‘, and no, it ain’t because everyone’s buying fancy graphics cards to play “Call of Duty” or mine digital gold. Nope, this time around, it’s about AI.

In the high-stakes race of tech bigwigs adding AI tools to their gear, Nvidia’s the guy selling sneakers. Google and Microsoft, among others, are making it rain with Nvidia chips for their AI ambitions. And boy, has that been good for Nvidia’s bank account. We’re talking about raking in more than $2 billion in profit over just three months.

Of course, they didn’t start by selling AI accelerators. Nah, they were all about gaming and crypto mining GPUs during the early pandemic days. But as that ship started to sink in 2022, Nvidia CEO, Jensen Huang, played his cards right, betting big on the data center boom. And wouldn’t you know it, his gamble paid off.

Their latest show and tell, Nvidia’s Computex 2023 keynote, was chock-full of AI goodies. They showcased games that can understand and respond to you like a human buddy, thanks to their Avatar Cloud Engine. And a new supercomputer that’s got more horsepower for AI than you can shake a stick at.

With the stock opening at over $400 per share, Nvidia now rubs shoulders with tech giants like Apple and Microsoft in the trillion-dollar club. That’s some rarefied air right there, folks. Only Amazon and Google share that space, while Meta used to, but fell from grace. Last week, Nvidia’s stock did a 25 percent hop, skip, and jump, and come Tuesday, it rose another 4 percent. Ain’t that somethin’?


As crypto embraces A.I., a major exchange scraps ChatGPT integration because ‘it’s very dangerous’

Looks like the bright boys and girls at Bitget, a big-shot crypto exchange, took a shot at riding the A.I. wave but ended up wiping out. They tried to stick OpenAI’s ChatGPT into their customer service department, hoping it would save the day answering all those customer questions. Turns out, it was like asking your grandma about the latest TikTok trends. The A.I.’s last update was in September 2021, so it was spreading old news, even recommended a crypto that had bit the dust.

Bitget users were about as happy as a cat in a bath, with 80% of ’em having a bad experience. After a fortnight, Bitget yanked the plug, with Gracy Chen, Bitget’s boss, saying that leaning too hard on A.I. can lead to some lazy decision-making. Still, they’re not about to throw the baby out with the bathwater – they plan to keep tinkering with A.I., reckon it could shake things up in the crypto world like DeFi did back in 2020.

So, here’s a thought: maybe A.I.’s still got a place in crypto, just not the one that folks first thought of. As always, it’s about striking that balance between shiny new tech and good ol’ human noggin.


Deepfaking it: America’s 2024 election collides with AI boom

Seems ol’ Hilldog and Joe Biden ain’t exactly what they appear to be in some viral videos. Nope, they’re what the tech whizzes call “deepfakes”. That’s a fancy term for videos so realistic, you’d swear they were real. Only they ain’t.

These tech prodigies have been training their computer gizmos on loads of online footage to make the fakes. Making a deepfake used to cost a bundle, but nowadays, it’s cheaper than a fancy cup of joe. As a result, you’ve got more of these phony vids than mosquitoes at a summer picnic.

The catch? Well, imagine your uncle Earl’s favorite conspiracy theory, but this time he’s got video ‘proof’. That’s right, deepfakes could muddy the waters between fact and fiction. Makes ya wonder if we’re all just gonna end up bamboozled, right?

Bigwigs at OpenAI are sweatin’ bullets over this. They’re like the head honchos in the world of AI, and even they don’t know how to keep these deepfakes in check. In the meantime, you’ve got some startups churning out AI tools like a factory assembly line, with fewer safety features than a Pinto.

As for the politicos, it’s kinda like watching the cats playing with Pandora’s box. You’ve got Trump sharing deepfakes on his social media and the Republican National Committee rolling out a political ad completely made by AI. Even the small fish in rural Michigan are getting in on the AI game, hoping to even the odds against the big dogs.

So, the moral of the story? Keep a keen eye out. Don’t believe everything you see on the internet. And remember, when it comes to these deepfakes, we’re all playing catch-up. Buckle in for the wild ride, folks.


Canadian AI computing startup Tenstorrent and LG partner to build chips

A Canadian AI startup called Tenstorrent, run by this former Apple and Tesla whiz kid, Jim Keller, just teamed up with South Korea’s LG Electronics. They’re about to churn out chips like a high-roller at a Vegas casino. Only these chips are the kind that juice up smart TVs, fancy car gizmos, and data centers.

Tenstorrent, already a billion-dollar big shot, has been quietly doing its thing since 2016. They craft computers that train and run AI models and dabble in both hardware and software. Keller, who’s famous for his work on chip design, jumped in the captain’s chair this year.

Now, LG’s first order of business is to use Tenstorrent’s AI chip blueprint to create its own chips. But this ain’t just about swapping blueprints, folks. There’s talk of Tenstorrent eyeballing some of LG’s tech for their own gadgets or maybe even for future customers.

Oh, and here’s a twist for ya. Tenstorrent’s got a chip in the works based on something called RISC-V. It’s an up-and-coming chip design that’s duking it out with the big dog, Arm architecture. Unlike most chip startups that stick to one lane, Keller’s crew is juggling both the AI chip and this processor. They reckon these two have to be two peas in a pod to keep up with the AI whirlwind.

Now, according to Keller, we’re still in the early days of this AI rodeo. But, in his words, folks have learned a ton in the last five years, and they’re making strides. So buckle up, folks, ’cause it’s shaping up to be one heck of a ride in the chip world.


What’s new in robots? An AI-powered humanoid machine that writes poems

Meet Ameca, a French-speaking, Chinese-speaking, poem-writing, cat-sketching robot with a rubbery blue face and a smile that’s all her own. Powered by generative artificial intelligence, she’s designed to chat, interact, and probably dazzle you with her talents.

This ain’t just any old robot, folks. Ameca was strutting her stuff at the International Conference on Robotics and Automation, the big kahuna of robot events, held in London. Picture it: robot cooking contests, autonomous driving challenges, brainy academics sharing their research, and startups flaunting their newest tech. It’s a bit like the Olympics for robots.

Amidst all this tech wizardry, there were also words of caution. Some of the biggest names in tech, including execs from Microsoft and Google, are sounding the alarm bells about the potential dangers AI could pose to mankind. They’re arguing that we need to put some serious thought into how we can lessen the risks of AI-induced extinction. Yes, you heard that right, extinction.

Meanwhile, the conference floor was a real spectacle. There were robot dogs running around, people using VR headsets to operate androids on wheels, and students from the University of Bonn showing off an avatar system that lets you control robotic hands. This system is so intuitive that anyone can get the hang of it in about half an hour.

One of the standout features of the event was the incorporation of AI systems into the mix. There’s a lot of buzz about blending AI like ChatGPT with robotics, which could open up a world of possibilities. Imagine being able to instruct a robot using natural language, no programming necessary.

Ameca is the creation of a British company called Engineered Arts. They specialize in robots designed for human interaction, perfect for roles like amusement park guides. According to Will Jackson, the company’s director, the biggest challenge for robotics these days is mechanical engineering, as AI has advanced leaps and bounds.

Ameca herself uses an AI image generator called Stable Diffusion for her drawing skills, and OpenAI’s GPT-3 for her quick-witted responses. When asked to compose a poem, Ameca came up with a few verses in a matter of seconds, paying homage to the Associated Press. Now that’s a robot with a flair for creativity!