- NATURAL 20
- Posts
- Secret Teams Making Google & Nvidia's AI Unbreakable
Secret Teams Making Google & Nvidia's AI Unbreakable
Red Team Hackers Crucial Role in Enhancing AI Security Protocols and Uplifting Industry Standards.
Wes Roth
September 04, 2023
Today:
Tech Giants Have Hired Red Team Hackers To Break Their AI Models
Tech giants like Google, Microsoft, and Nvidia are hiring "red team" hackers to poke holes in their AI systems. These hackers try to break the AI to find weak spots that need fixing. They've even tested stuff like making AI spit out harmful or incorrect info, just to see if it would. It's a balancing act. Make the AI too safe and it's boring; let it roam free and it might cross some lines.
As AI gets more complex, these red teams are key to keeping things on the up-and-up. They're not just looking for obvious glitches, but are also trying to stop the AI from leaking personal info or getting tricked by bad data. With a lot of eyes watching how safe and reliable AI is, these teams also give tech companies a chance to say, "Hey, our AI is the safest around."
IBM and Salesforce Team Up To Help Businesses Accelerate Adoption of Trustworthy AI
IBM and Salesforce are joining forces to help businesses get on the fast track with AI, particularly in customer relations (CRM). IBM's Consulting team is bringing their know-how to help businesses implement Salesforce's AI tools like Einstein and Slack smoothly. They're also mixing in their own Watson tech to help make sense of data and improve the user experience.
The goal? Make workers more productive and customer service top-notch. They're not just talking the talk; IBM has even used Salesforce's tech for its own upgrades. Both giants believe this partnership will shake up the way companies engage with their customers and help them get ahead in a competitive world.
Korean internet giant Naver explores robotics, AI and autonomous driving
South Korea's Naver, often called the "Google of South Korea," is diving deep into robotics, AI, and self-driving tech. Inside their snazzy new HQ, dubbed Naver 1784, they've got 100 robots called Rookies zipping around delivering coffee and parcels. These robots are getting their brains upgraded in a new data center this November. On top of that, Naver owns a bunch of other companies doing everything from e-commerce to digital comics. They've also recently launched a fancy AI service, showing they're not kidding around in the tech race.
Naver has been toying with robots and AI since 2013 and even bought a French research center to help out. They think robots could be the next big thing, replacing mobile phones someday. Their vision? Robots everywhere: homes, hospitals, malls, you name it. While they're not there yet, they're using the tech they develop to keep improving their in-house robots.
Ads for AI sex workers are flooding Instagram and TikTok
AI chatbots are getting raunchy, and social media platforms are getting swamped. Despite Facebook, Instagram, and TikTok's past efforts to block steamy content, a new wave of ads for AI chatbots is sliding under their radars. These aren't just any ads; they're promoting barely dressed, flirty bots.
Tech startups are throwing these ads all over social media, luring folks with "NSFW pics" and uncensored chats, some even using kid-favorite characters like SpongeBob to market their apps. NBC News did some digging and found 35 app developers pushing these racy ads on Meta-owned platforms. Another 14 were doing the same song and dance on TikTok. Both companies yanked some ads after they got wind of NBC's research, but it's unclear why these AI ads got a pass while real humans have historically been shut down on these platforms.
Dell shares hit record high after report, forecasts impress with AI in mix
Dell's stocks are soaring, hitting an all-time high after some pretty solid financial projections, partly thanks to artificial intelligence (AI) and people buying more computer stuff again after a bit of a lull. The share price topped at $70.28, which is way up from earlier in the year.
Experts are saying the demand for business tech is on the upswing. Dell's latest numbers are better than what most folks predicted, especially in the department where they sell servers that work well with AI. Even though not everyone sees AI as the main reason for this uptick, it's clear that AI's making a difference.
After the report, a bunch of financial pros upped their price guesses for Dell. Some big banks like Wells Fargo and Citigroup are even more optimistic about where Dell's stock price might go.
Hyderabad Firm Unveils India's First AI-Powered Anti-Drone System
A company in Hyderabad just rolled out India's first-ever AI-driven drone defense system, called Indrajaal. This bad boy isn't just fancy; it can protect everything from nuclear spots to entire cities from any drone. This big reveal comes from Grene Robotics, a company that’s deep into AI security. Indrajaal stands out because it's like a one-stop-shop to protect against moving threats, kinda like having a security guard who can stop anything coming your way.
This high-tech defense tool got a thumbs up from Uttarakhand Governor Lieutenant General Gurmit Singh. He remembers when drones attacked Jammu airport back in 2021 and sees Indrajaal as India's answer to such threats.
The founder, Kiran Raju, mentions that this system is like a Swiss army knife with 12 different tools, all juiced up by AI. It can spot, track, and stop threats in almost no time. It's ready to protect a huge chunk of land (about 4,000 square kilometers) from everything: drones, smart bombs, and more.
A developer built a 'propaganda machine' using OpenAI tech to highlight the dangers of mass-produced AI disinformation
A developer whipped up an AI "fake news factory" using OpenAI tools like ChatGPT in just two months for under $400 a month. Called CounterCloud, this project showcased how a breeze it is to churn out mass propaganda. The brain behind this? An anonymous someone going by Nea Paw, supposedly in cybersecurity.
Through a YouTube vid, Paw broke down the process, revealing how the AI could spit out phony stories and even rewrite real ones to sow doubt. They added bells and whistles like voiceovers and fake journo profiles for that "real feel". While the system's cranking out believable stuff round the clock, Paw's holding back on releasing it online – because who knows where that wild ride would end.
Floworks is building an AI assistant to change how workers interact with software
Floworks is cooking up an AI buddy to make office software easier to handle. Started by Sudipta Biswas and Sarthak Shrivastava after some wise words from big-shot investor Vinod Khosla, this AI helps you with boring office jobs. Just chat with it, like "Hey, add this person to my contacts," and it’ll take care of it using apps like Gmail or Salesforce.
Right now, it's like a helper for sales folks, but the plan is to help out in other job areas too. You chat with this assistant in Slack, and it even asks you questions if it needs to know more. They're looking to add Microsoft Teams support down the line. They've got a test version out, making around $250k, and aim to grow by teaming up with partners. Their unique twist? This AI can chat with tons of apps, not just a select few.
Based in San Francisco and India, they're all about diversity, thinking it's the secret sauce for a winning product. They just bagged $1.5 million from big names like Y Combinator and Sense AI. Cool, right?
AI Cracks the Code on Odor Perception
Scientists teamed up with AI pros to crack the mystery of how smells work. They used machine learning to predict how a chemical will smell to humans. This AI beat the human sniff test over half the time! It even threw in a curveball by doing something it wasn't trained for—gauging how strong a smell is.
This isn't just about explaining a weird science question; it could change the game for making perfumes, flavors, or even bug repellents. The AI got trained on 5,000 known smelly chemicals and even helped scientists make a new kind of map that shows how smells relate to each other. So, this tech might just put our noses out of business!
AI-powered BeFake is a real app, not a BeReal parody…and it has $3M in funding
BeFake, this new app playing off of BeReal, lets you jazz up your photos with AI. Instead of the usual "look at me now" photo, it lets you snap pics and then use AI to change 'em up. You can use their options or come up with your own, like making it look like you're on top of Mount Everest. It's from the brains behind a big gaming company, Machine Zone, which made over a billion bucks from games. They got the idea from gamers connecting online with fake identities.
Even though BeFake just popped up on the App Store and Google Play, it's growing like crazy. They make money by charging for more AI features, but they do give users a chance to use the app for free every day.
These folks know they gotta stay on their toes to keep users interested, so they're all about quick updates and listening to what users want. They've got a cool $3 million in the bank from investors, including big names like Peter Thiel.
The USAF Pairs Piloted Jets With AI Drones
The U.S. Air Force is teaming up human pilots with AI drones, testing out if the combo can kick some serious tail in the sky. Think of it like a buddy cop movie, but with F-22 Raptors and AI drones called Valkyries. This ain't cheap—Uncle Sam is asking Congress for $5.8 billion to really get the ball rolling.
A company called Kratos is making these Valkyrie drones and says they're the best bang for your buck. A single Valkyrie could cost as low as $2 million, way less than its rivals. These drones aren't slouches either; they're fast, can go long distances, and can even carry other mini-drones.
AI Took the Stage at the World's Largest Arts Festival. Here's What Happened
AI took the stage at Edinburgh Fringe, the world's biggest arts fest, and it was a mixed bag. The big showstopper? A "sex robot" that turned out to be a comedian in disguise. The gig got folks talking about how AI and humans mix in the arts, especially with workers worried that tech like ChatGPT could swipe their jobs.
Improv shows had robots helping human performers, but mostly people were laughing at the AI, not with it. One standup comic even had her own heckling robot. Yeah, it's clear we've got questions about AI — like, can it be creative without humans steering the ship? Artists still bring something to the table that machines can't match: emotion and that human touch. But as tech improves, we're all watching to see how this odd couple — artists and AI — evolves.
What'd you think of today's edition? |
What are MOST interested in learning about AI?What stories or resources will be most interesting for you to hear about? |
Reply