- NATURAL 20
- Posts
- Coca-Cola® Y3000's AI-Inspired Taste!
Coca-Cola® Y3000's AI-Inspired Taste!
Groundbreaking partnership with artificial intelligence to present Coca-Cola® Y3000—a zero-sugar beverage of a future world.
Wes Roth
September 13, 2023
Today:
Coca‑Cola® Creations Imagines Year 3000 With New Futuristic Flavor and AI-Powered Experience
Coca-Cola's taking a leap into the future with a new drink flavor and a cool AI twist. They mixed ideas from fans worldwide and some fancy computer smarts to come up with Coca-Cola Y3000 Zero Sugar. The goal? Figure out what a Coke from the year 3000 might taste like.
They hope this drink stays as refreshing and hip a thousand years from now as the regular Coke is today. You can snag this zero-sugar drink for a short time in spots like the US, Canada, China, Europe, and Africa. Folks in the US, Canada, and Mexico can also try an original flavor version.
The drink's design is all space-age vibes, with changing colors like violet and magenta on a shiny silver backdrop. Even the classic Coke writing's got a modern twist to fit the futuristic theme.
Salesforce embeds conversational AI across the platform with Einstein Copilot
Salesforce is upping its AI game with a new feature called Einstein Copilot. Introduced at their Dreamforce conference, this tool lets users ask questions in everyday language to pull up info they need, without jumping through hoops or clicking a bunch of times. It's an extension of their older Einstein GPT feature, which helped with automated tasks like answering customer queries or drafting emails. This new add-on aims to make workers more efficient and improve customer interactions across different areas like sales and e-commerce.
What sets Salesforce apart? They've got data from many customer touchpoints, giving them an edge in delivering more useful and accurate AI responses. But Salesforce knows these AI systems aren't perfect, so they're adding a "trust layer" for better security and data governance. While Einstein Copilot is still in beta testing, the trust layer will be available soon.
World scale inverse reinforcement learning in Google Maps
Google Maps has upped its game by making its routing even smarter. The tech giant used something called inverse reinforcement learning (or IRL for short, but let's not get bogged down) to study real-world driving patterns and figure out what routes people actually prefer. The big challenge? Doing this on a global scale without frying their computers.
So, Google teamed up with DeepMind to make the math manageable. Their new algorithm, called RHIP, fine-tunes the trade-offs between things like time, tolls, and road conditions. The result? Routes suggested by Google Maps are now 16–24% more likely to match the routes people actually take. Plus, it even helps with eco-friendly travel options like scooters. The cherry on top? The system's faster and cheaper to run, which means everyone wins.
Kakao plans to sell ownership stake in Starship Entertainment to raise funds for AI research and development
Kakao's about to sell its piece of Starship Entertainment to bankroll more AI work. Why? They're trying to outpace their big competitor, Naver, in the AI game. Last year, Kakao chucked about $754 million into AI, but that's peanuts compared to what Naver spent.
Kakao hopes this year's sell-off will add a cool $75.4 million to their funds. Keep an eye out this October, Kakao's dropping a new Korean language AI tool. Oh, and if you're wondering, Starship's got some big music names like IVE and MONSTA X.
With ‘GitHub for data,’ Gable.ai wants to connect software engineers and ML developers
Seattle startup Gable.ai just got $7 million in seed funding to be the "GitHub for data." They're aiming to fix a big ol' headache in the tech world: making sure the data feeding into AI applications is top-notch and trustworthy.
The team behind Gable.ai previously led the data department at Convoy, a $4 billion digital freight company, so they know firsthand how messy and unreliable data can wreck even the best AI models. Their platform helps data makers and users collaborate smoothly, preventing costly hiccups in the data flow.
Big-name investors and other data companies are already jumping on board, so it looks like Gable.ai is filling a gap that's been a real pain point for both software engineers and machine learning developers.
Druid, a conversational AI platform for enterprises that integrates with ChatGPT, raises $30M
Druid, a Romanian AI startup, just scored $30 million in new funding to grow its business in the U.S., where it's already making most of its money. The company uses conversational AI to help other businesses automate stuff like customer support. What sets it apart? It's got a partnership with Microsoft and integrates with ChatGPT, making its virtual assistants smarter and better at understanding what customers are asking.
While not a one-size-fits-all solution, Druid's tech can be a big help to companies looking to up their customer service game. They plan to use the new cash to expand globally, especially in the U.S., and are even thinking about moving their main office to Austin, Texas.
Hailo Expands Hailo-8 Accelerator Lineup for Entry-Level Edge AI
Israeli startup Hailo is beefing up its AI chip game, targeting everything from self-driving cars to city cameras. They've rolled out an improved Hailo-8 chip that's super-fast but doesn't guzzle energy. Basically, they've reorganized how data flows in the chip, making it more efficient. They're even saying their chip outperforms similar products from Nvidia, a big name in the tech world.
The company has also expanded its PCI cards—those are plug-ins that make your computer more powerful—for even heftier tasks like video management. All of their stuff is open-source, so you're not stuck with one vendor.
Enfabrica, which builds networking hardware to drive AI workloads, raises $125M
Enfabrica, a tech startup making special computer chips to help AI run smoothly, has scored big with $125 million in new funding. Their total funding now sits at $148 million. Led by big names like Nvidia, the new money will help the company expand its team and refine its products. Despite a tough market for raising funds, especially in the techy chip world, Enfabrica stands out. Why? Because as AI keeps growing, the need for handling all the data and calculations is becoming a major headache.
The company was co-founded by Rochan Sankar and Shrijeet Mukherjee, pros who used to work at Broadcom and Google. They realized that current computer networking hardware just can’t keep up with the heavy lifting required by modern AI systems. So they've been building their own version that's way better at moving tons of data super fast.
FinTech firm Hadrius nets $2m seed round for AI-powered SEC compliance
Hadrius, a FinTech startup focused on making SEC rules easier to follow using AI, just bagged $2 million in a seed investment round. Some big names like Y Combinator and Lynett Capital threw their money into the pot. This AI system they've got? It’s all about simplifying a bunch of tedious SEC tasks for companies. With this fresh cash, Hadrius plans to boost their reach and spruce up their tech.
The brains behind Hadrius? Thomas Stewart, Som Mohapatra, and Allen Calderwood. These guys aren't rookies; they previously started the robo-advisor, Quantbase. Som, a bigwig at Hadrius, thinks the old ways of managing SEC stuff are outdated and a headache. He's convinced Hadrius is the game-changer many firms need. And given the caliber of their tech team - with folks who used to be with companies like Google and Amazon - they might be onto something big. This $2 million is just another feather in their cap as they march forward.
AI Miracle: How ChatGPT Cracked A Boy’s Chronic Pain Mystery
Young Alex had a pain mystery that left 17 docs scratching their heads. For three long years, his once lively spirit got shadowed by tantrums and unexplained aches. Even a trip to the ER didn't clear things up. His mom, Courtney, tried everything - from ortho to physio. No luck. Desperate, she tossed her problem into ChatGPT's lap, an AI chat program.
After pouring in MRI notes and symptoms, ChatGPT played detective and pointed at "tethered cord syndrome." In plain words? Tissue was tying down Alex's spinal cord, causing all sorts of problems. With that lead, Courtney found a specialist who confirmed the AI's guess. Alex had surgery, and now he's on the road to feeling like himself again.
Through this rollercoaster, technology and a mom's never-give-up spirit teamed up, showing that sometimes AI and humans can mix to cook up something amazing. So, for those worried about tech taking over, remember: it's a tool, and in the right hands, it can work wonders.
8 More Companies Pledge to Make A.I. Safe, White House Says
Eight big-name companies, including Adobe and IBM, have promised to play it safe with A.I., the White House announced. They're hopping on board with giants like Amazon and Google who made similar promises earlier. Though these commitments aren't government rules, they're about testing the tech and sharing info about any risks.
The push for these safety moves comes after A.I., like the ChatGPT chatbot, started raising eyebrows with concerns about jobs, spreading false info, and getting too smart for its own good. However, some groups aren't too thrilled about these tech giants having such a big say in A.I. rules. They feel these companies have too much power and shouldn't overshadow the little guys.
Elon Musk and Mark Zuckerberg will talk AI rules in Senate forum on Wednesday
Big tech honchos like Elon Musk and Mark Zuckerberg are meeting up in Washington this Wednesday. They're talking to lawmakers about how to handle the rules around AI, or artificial intelligence. Senate Majority Leader Chuck Schumer hopes this meeting will help set the stage for future laws on AI. Some other CEOs and industry bigwigs were also invited, but Amazon said "no thanks" due to scheduling issues.
There's a mix of hope and eye-rolling about the summit. Some folks think it's all talk and no action, especially because smaller tech companies weren't invited. They're the ones really driving AI innovations, not just the giant companies. Some are also worried about repeating past mistakes, like not regulating social media soon enough.
Companies That Use AI To Help You Cheat At School Are Thriving On TikTok And Meta
AI companies that help students cheat are blowing up on TikTok and Meta. Even though these bots like ChatGPT can help with schoolwork, they're not perfect and often spit out wrong info. Some businesses even mix AI with real people to make cheating hard to spot.
These companies are advertising big-time on TikTok and Meta, even though it's illegal in places like England and Australia. A professor, Michael Veale, found out about this when checking ads on these platforms due to EU rules. 11 AI cheating services were found. TikTok says they've taken down the bad ads and banned the accounts behind them. Meta hasn’t said anything.
But here's the thing: the rules around these ads are fuzzy. Some think even normal tools like autocorrect could be seen as cheat tools! With AI becoming a regular thing, experts like Thomas Lancaster think it's gonna be super hard to stop these cheat ads since it's tough to spot who's using AI.
AI-Powered ‘Thought Decoders’ Won’t Just Read Your Mind — They’ll Change It
Alright, here's the lowdown: Scientists are making headway in decoding human thoughts using fancy brain scans and AI. Imagine wearing a thinking cap that reads your mind, translating your thoughts into words. Sounds like sci-fi, right? They've already done stuff like reconstructing a Pink Floyd song someone was listening to. However, you gotta keep your butt in a scanner for hours, so it ain't perfect yet.
Our thoughts aren't just locked boxes inside our heads; they're shaped by everything around us. So, these thought-decoding machines won't just read our minds; they could also change how we think. Like, imagine if an algorithm not only knew you liked burgers but also made you crave them.
AI model speeds up high-resolution computer vision
Researchers from MIT and MIT-IBM Watson AI Lab have cooked up a new computer vision model that’s faster and more efficient than the old ones. This tech helps self-driving cars see and react to things, like a speeding cyclist or a parked truck, by analyzing every tiny part of a detailed image.
The usual models were too slow, especially for gadgets like phones. But the new MIT model is up to 9 times faster and just as accurate, if not more. Imagine a car computer quickly recognizing objects in real time, making roads safer. This could also help doctors with medical images or improve video games.
Experts from big tech companies like AMD and Oracle are buzzing about it, highlighting how this might change the game in AI. The MIT team wants to keep pushing the boundaries with their creation.
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