States are Swiping Right on AI—Here's How They're Making Sure It's Relationship Material, Not a Total Creep!

States are waking up to the fact that artificial intelligence is not some sci-fi fantasy. It's here, and it’s like a powerful new roommate that you have to figure out how to live with. You know, will they clean their dishes or turn your life into an Orwellian nightmare? Some states are saying, "Before we hand over the keys, let's get to know you."

Minnesota is doing its homework like a student on an Adderall binge. They're diving deep into how AI could potentially play Big Brother with civil rights, especially when it comes to law enforcement. Then there's North Dakota, not wanting to be left in the tech dust, examining how AI might throw its hat in the ring for matters like job markets and even the 2024 elections. That's right, folks, Skynet could be a swing state.

Three amigos—Connecticut, Texas, and Washington—are forming what I'd like to call the Justice League of AI Panels. Their mission? To check if automated systems are pulling a fast one and discriminating against citizens based on race, gender, or religion. I mean, it's 2023; if my Roomba can avoid dog toys, we can program fairness into our algorithms, right?

Wisconsin also joined the club this week, with Governor Tony Evers issuing an executive order for an AI task force that sounds like it was named by a committee: "The AI Task Force Devoted to Labor Issues." Seriously, I've seen Twitter handles with more charisma.

Here's where it gets spicy. Lawmakers aren't just looking to put guidelines around state operations. Oh no, they're setting the stage for regulating the tech world, too—everything from algorithmic decision-making to those fancy language models that process data faster than I can say "pass the sriracha."

Connecticut Sen. James Maroney likens lawmaking to painting a house. No, really. He says task forces and research are the "prep work." Kinda like taping up the trim before going to town with that roller. It's a little less sexy but ensures you don't end up with a polka-dot living room. If you skip the prep, don't blame anyone if the end result is a messy wall of bad laws and unintended consequences.

The brass tacks? States are shaking off their hesitancy about blue-ribbon panels because, let's face it, the stakes are too high. We're not just talking about your Spotify playlists here, but issues that affect "liberty, finances, livelihood, and privacy interests." Essentially, states are saying, "We don’t want to mess this up and end up in a 'wild, wild west' of unregulated tech."

Vermont, ever the overachiever, is already schooling us on how it's done. They've created a public inventory of the state's AI assets and even set up an Artificial Intelligence Advisory Council. Vermont is the friend who not only brings a tent to the camping trip but also remembers the bug spray and portable phone charger.

What’s the takeaway? If you're not thinking about how AI will affect your life, start now because the states are gearing up to make some real decisions. And remember, we're all in this roommate selection process together. Let's just make sure AI is more of a helpful, dishwashing type rather than the one stealing our privacy... or worse, our snacks.

From Free Cruises to Fake CEOs: How Deepfake Scams Are Fooling Even the Tech-Savviest Among Us

Ah, remember the good ol' days when scam calls were just robotic voices claiming you'd won a "free cruise"? Well, welcome to Scamming 2.0, where hackers use deepfake tech to sound exactly like your boss, asking you to wire money ASAP.

Take the CEO of a British energy company, for instance. A couple of years ago, he got hoodwinked into sending €220,000 ($249,000 for those of us who don't eat croissants) to a scammer who mimicked the voice of his German counterpart. Euler Hermes, the insurer, confirmed this as an AI-generated voice scam. I mean, if you can't trust your ears, what's left? Your gut instinct that always tells you to buy crypto right before it tanks?

As if voice deepfakes weren't enough, video deepfakes have entered the chat. Imagine jumping on a Zoom call and seeing what appears to be your boss. And this isn't some badly dubbed Kung Fu movie; we're talking AI-crafted, real-time deception. Mandiant, a cybersecurity firm, recently documented deepfake tech specifically designed for phishing. The cost? A measly $20 a minute. That's cheaper than my last therapy session talking about my trust issues, which are clearly well-founded.

But wait, it gets crazier. Udi Mokady, the big kahuna at CyberArk Software, was gobsmacked to find himself staring at his own deepfake in a Microsoft Teams call. The culprit? His employee, Gal Zror, who did it as a "Hey, this could happen to us" warning. Zror didn't just Google "how to make a deepfake;" the guy deep-dived into forums frequented by, get this, deepfake pornographers. The point? To show that if he could create a deepfake, anyone could.

Zror even presented his deepfake magic trick at a hacker conference, earning himself applause as he turned into the conference organizer, right there on screen. But let's face it, CyberArk is a small fry compared to giants like IBM or Walmart. Imagine tricking employee number 15,003 who might never even have met their CEO. That’s more dangerous than grandma mixing up salt with sugar in her apple pie recipe.

So, what's the moral of this Black Mirror episode we're living in? First, educate your team, especially if your team is the size of a small country. Second, add layers of security like it's a winter coat—two-factor authentication is your friend here. And finally, trust but verify. If you get a call from "me" asking you to invest in a llama farm, maybe shoot me a text first, eh? Unless, of course, you know that investing in llama farms has been my dream since 3rd grade. Then it's totally legit. 🦙 

Top Gun Meets Terminator: Why the Air Force's New AI-Powered Drones Are Both Mind-Blowing and Terrifying

We're about to talk Top Gun, Skynet, and billions of your tax dollars, all swirling together in an aerial ballet choreographed by Uncle Sam and Silicon Valley. Yep, the U.S. Air Force is eyeing a hefty budget to build 1,000 to 2,000 AI-powered drone sidekicks. Think of these XQ-58A Valkyrie aircraft as Goose to your Maverick, if Goose was made of metal and had an affinity for "suicide missions."

We're not talking about some far-off fantasy; these bad boys are already in the works. Later this year, Valkyrie drones will be put to the test, simulating their own hunt-and-destroy missions over the Gulf of Mexico. "Talk to me, Goose!" Except Goose is an AI algorithm optimizing kill vectors—kinda cool but also kinda terrifying, right?

These AI pilots aren't slouches, either. The Valkyrie drones can zoom around at 550 mph and soar at 45,000 feet, all while hitting a range of 3,000 nautical miles. And get this: each one's price tag could be as "low" as $3 million. Compared to a manned jet, that's like choosing between a Ferrari and a used Honda Civic. The Air Force is pretty serious about these robot wingmen; they're asking for a cool $5.8 billion over the next five years.

I remember playing with remote-control airplanes as a kid and thinking, "Wouldn't it be cool if this could fly itself?" Well, I got my wish, but the grown-up version is equipped with military-grade firepower and has been beta-tested alongside F-22s and F-35s. It's like your childhood fantasies got a Ph.D. in Aeronautical Destruction.

But hold your horses, Sarah Connor. There's a Skynet-shaped cloud hanging over this techno-utopia. Some folks are really, really worried that we're stepping into a "Terminator"-esque nightmare. Human rights advocates are sounding the alarm, saying we're crossing a moral Rubicon by letting algorithms make life-and-death decisions. It’s one thing for AI to recommend your next Netflix binge; it’s another for it to decide who gets a missile up the tailpipe.

Even the United Nations is throwing shade. Secretary-General António Guterres has been on the record since 2019, saying these killbots are "politically unacceptable" and "morally repugnant." Ouch. That's not the kind of review you want on your Tinder profile.

So, what's my take? Proceed with caution. Tech like this could revolutionize warfare, save human lives on our side, and make military operations more efficient. But if we rush in without clear guidelines, we're flirting with chaos and a bucket load of ethical landmines. You don't have to be John Connor to know that handing over the kill switch to a machine should make us all pause and think.

From Cyber Hero to Possible Villain: How AI Could Either Save or Sabotage Your Online Security

Ah, AI: the buzzy tech that's everywhere, from making your Spotify playlist to, uh, helping me write this sentence. Yeah, it’s literally writing the story about itself; how meta is that? But the conversation’s been heating up about AI and Large Language Models (LLMs) like ChatGPT potentially flipping the cybersecurity game on its head. No kidding. It's like assigning your puppy to guard the house. Cute and helpful, but there's a good chance things could get messy.

Now, don’t get me wrong. AI could be the fairy godparent cybersecurity has been waiting for—magically drafting and scanning code, real-time threat analysis, and relieving overworked, under-caffeinated security experts from the grunt work. Imagine swapping those mind-numbing Excel sheets for some good ol' human problem-solving. More time for company ping pong tournaments, anyone?

But let's pump the brakes. AI and LLMs are still the awkward teenagers of the tech world. They’ve got potential, but they’re also a liability. Turns out, the same tech that could boost cybersecurity can be co-opted by the baddies, the online villains. Picture a cat burglar using your own ladder to sneak in through the second-floor window. Not so convenient now, is it?

Now, I hear you asking, “But how bad could it be?” Well, get this: Malicious types are already using AI to spin off new variants of harmful code—imagine replicating a webshell attack like it’s going out of style. And trust me, they’re gonna keep iterating faster than you can say "zero-day vulnerability."

And if you thought that open-source is the magic potion, think again. A lot of the baddies are using open-source LLMs to automate their evil deeds. It's like offering a free getaway car with a map to the bank vault. Before we know it, we'll see a boom in zero-day attacks, and that’s a day no one’s looking forward to.

So what's the playbook here? No silver bullets, but how about fighting fire with fire? Just like the bad guys, organizations should use AI to scan their code for security holes faster than you can find plot holes in a soap opera. If you’re pulling in code from the cyber realm’s equivalent of a public library, make sure you’re not also importing its malware history.

The key takeaway? Whether you're an Elon Musk fanboy or think AI is the beginning of Skynet, this tech isn’t going back in the bottle. It’s like learning to live with a roommate who sometimes eats your food. You've got to set boundaries. And folks, the industry bigwigs are echoing the same thing. An "AI pause" might be on the horizon, but for now, let’s be smart about this tool—before it's smart enough to be dumb about us.