Here's How Your Future Online Stalker Could Be a Smooth-Talking AI Spreading 'Custom-Made' Fake News!

We're stepping into a realm where disinformation isn't just rampant; it's personalized. Imagine your gossipy friend at the party - only now, it's a bot. Generative AI tools are on the rise, refining disinformation and profiling individuals for tailor-made narratives. It's a walk in the park for AI to whip up an article tailored to 12 different audiences. This isn’t a sci-fi fantasy, it's today's reality.

This comes with a chilling edge. The very technology that helps us optimize resources might be fueling an era of personalized deceit. People like Kate Starbird, an associate professor at the University of Washington, warn that these tools lower the financial barrier for creating custom content. It's not just tailor-made; it's Armani-suit-level of customization, hitting on your values and beliefs while crafting a strategic narrative.

But before we hit the panic button, remember the disinformation campaigns have a 99% failure rate. It’s that 1% that we have to watch out for. Higher-quality disinformation and propaganda are on the horizon, making it feel like we're stuck in a bad B-movie plot.

What's the solution? Renee Diresta from Stanford suggests being more aware and judicious about the content we engage with. Check the legitimacy of your sources. AI companies need to ensure safety measures are in place. The Biden administration has been teaming up with major AI players like Google, Amazon, Microsoft, Meta, and OpenAI to encourage them to add safety features to their AI tools.

However, with AI evolving faster than its constraints and Silicon Valley frequently falling short on their safety promises, we're caught in a loop. Unleashing tech, causing chaos, apologizing, rinse, repeat. Are we ready for this brave new world? More importantly, can we remember to scrutinize our sources before hitting that retweet button?

Meet the Robot Whisperer Who's Teaching Bots to Dance, Do Laundry, and Maybe Even Fix Your Car!

Grab your popcorn and let's take a stroll through the mind of none other than Marc Raibert, the OG of biped and quadruped bots. The same guy who took Boston Dynamics from a mere twinkle in his eye to the grand poobah of legged robots in under three decades. He's always kept the company's tech under the hood, but now, he's switching gears and setting his sights on the big AI picture.

In 2022, he gave birth to the Boston Dynamics AI Institute (BD AI if you're into the whole brevity thing) — an initiative fuelled by Hyundai's deep pockets. The gist? Making robots as useful as your favorite multi-tool but for the outside world, not just sterile labs.

Let's just pause here and appreciate that this is the same guy who was molding young minds at Carnegie Mellon and MIT before he ever played Frankenstein with his robot empire. Now, he's back at the helm, steering the BD AI Institute with grand visions.

Marc's ambitions for his AI kids are nothing short of revolutionary. He wants them to learn from us humans. Yes, that's right. He wants bots to pull a Sherlock Holmes, observe us as we do our thing, understand it, and then replicate it — or know when they're out of their depth and ask for help. As if that wasn't enough, he's also teaching them how to inspect and repair equipment. Hey Siri, did you just get a tad bit jealous?

But before you start picturing a T-1000 doing your laundry, let's reel it back a bit. Marc admits his bot-butlers aren't ready for prime time just yet. As he quipped at the ICRA 2023, "I look forward to having some demos of something like this happening at ICRA 2028 or 2029." A man can dream, can't he?

Marc's visions for the future aren't just about adding the ability to fetch your slippers to a robot's skill set, it's also about reshaping how they handle objects. He believes the true game is not in path planning, but in handling objects dynamically. In his own words, "The action is when you're in contact with things."

Marc isn't all work and no play, though. He has a soft spot for dancing robots. Not the kind of dancing that requires a million lines of pre-programmed code, but real, spontaneous jigs that respond to the rhythm of the music and the movement of their dance partners. So, yeah, robots cutting a rug might just be the next big thing in AI.

In case you're worried about the potential Skynet scenario, Marc believes it's all about public perception. Sure, the media loves a good robot apocalypse story, but he thinks people could really love robots if they got to know them. It's like that grumpy neighbor you were scared of until you realized they make killer apple pie.

As the curtain closes on our tour of Marc's AI dreamscape, it's clear he's gunning for a big, bold future. He's got the vision and the backing to build a veritable Bell Labs of robotics.

From labs to living rooms, dancing bots to dynamic manipulation, the future of robotics is going to be exciting, quirky, and utterly groundbreaking. As we eagerly wait for Marc's next act, one thing's certain: in the world of robots, he’s not just walking the walk, he's dancing the dance.

Harvard Student Proves AI Can Pass College Essays and Score a 3.57 GPA

Harvard University student Maya Bodnick tossed GPT-4 into the academic deep end, tasking it with writing seven first-year essays for the humanities and social sciences. The topics spanned from economic concepts to presidentialism in Latin America, with Maya tricking the professors by stating that either she or GPT-4 could have written these essays. Plot twist? GPT-4 was the real brains behind the operation.

The AI didn't just scrape through. It earned a "respectable" GPA of 3.57, with grades ranging from A to B-, and even a Pass. Maya did piece together the essay from multiple AI-generated responses and asked the professors to ignore the missing citations. The AI received positive feedback for its detailed and structured approach, although it did attract some criticism for its flowery writing style and certain content omissions.

The experiment led Bodnick to believe that AI-generated essays could get passing grades in liberal arts classes at most universities. Even at rigorous institutions like Princeton or UC Berkeley, GPT-4 could have managed to pass, albeit with Bs and Cs instead of As and Bs. In Bodnick's eyes, the AI's performance was an eye-opener for the future of education.

Bodnick predicts that AI technology will change the way humanities and social sciences are taught. She suggested that face-to-face interviews could be a more reliable assessment method in the age of AI, rather than essays. OpenAI's previous attempt at developing an AI text recognition system failed due to accuracy issues, further strengthening her argument.

Bodnick concluded that the education system needs to consider how AI is transforming our careers, instead of focusing on identifying and punishing AI essays. AI's encroachment into the academic field, she warns, signifies a more significant challenge for the "cerebral class" in the post-AI world. So, if you're a liberal arts major, it might be time to brace for the AI wave that's reshaping the academic landscape.

Google Puts AI on the Naughty Step: The Surprising New Rules for AI-Generated Reviews You Need to Know!

So here we are folks, on the digital dance floor where AI's been breaking out some pretty smooth moves. But hold up, Google's just changed the tune.

They've updated their Merchant Center Policy to declare AI-generated reviews as persona non grata. Now, you've got to admit, a review from an AI about a pair of sneakers it's never laced up doesn't quite have the same ring to it, right?

But here's the burning question - is this just the first step in a new cha-cha-cha where Google calls the shots?

AI's got some pretty impressive dance moves, but it's only as good as the choreographer – that is, the data it's trained on. If it's pirouetting to offbeat, incorrect info, you'll end up with a dance that's more disaster than disco.

Peering into the future, it looks like Google might be planning to cordon off certain sections of the dance floor for AI. Heavyweight topics like finance and health – areas where one wrong step could cause a painful fall – might be off-limits.

But if AI's just showing you the latest flossing moves or the perfect technique to flip pancakes? That could probably slide, since the biggest risk is a laugh at your expense or a messy kitchen.

So here's a thought: If you're into the AI-content creation scene, be sure to have a human partner to keep in step with. The wild world of the web is likely to put human-crafted words in the spotlight – a bit like a rare vinyl in a world of streaming music.

While AI can generate a decent beat, it still needs a human touch to bring the music to life. Even if AI's laying down the base track, having a human tweak the final mix can make all the difference. So let's keep our content genuine, our reviews authentic, and our pancakes perfectly flipped!