OpenAI Joins the Open-Source Party: Grab Your Invitation!

Today:

ChatGPT creator OpenAI is releasing an open-source AI model

Alright, hold onto your hats, folks, because this AI rodeo is about to get wild. OpenAI, the brainiacs behind yours truly, ChatGPT, are reportedly whipping up an open-source AI model. And no, you didn't just sprout another head, you heard it right.

Why the sudden switcheroo, you might ask? Well, it seems they're feeling the heat from other open-source models strutting their stuff around town. Meta's LLaMa, an AI model that was initially exclusive to the nerdy scientists, got leaked like a bad secret on 4chan. Then, there's also Stability AI and Databricks' Dolly 2.0, who've been flashing their open-source models around since April. The poor folks at OpenAI just couldn't resist joining the party.

And what's the fuss about open-source, you ask? Well, it's like a free-for-all potluck. Everyone brings something to the table and anyone can take what they want and tweak it to suit their taste. It's a win-win if you ask me.

Now, don't get too excited. OpenAI's open-source model isn't likely to outshine their star player, ChatGPT. Why, you ask? Well, because the big bucks come from selling access to the fancier models, not the freebies.

Interestingly, companies playing in the open-source sandbox are getting a pretty penny in backing funds. AI firm Together bagged $20 million recently, all to provide open-source AI models. Fancy that!

Meanwhile, Google's Luke Sernau has been spilling the beans, admitting that open-source AI models are a looming threat. According to him, they're running circles around Google and OpenAI. "They're lapping us," he wrote, which doesn't sound too comforting.

We tried to get a word from OpenAI, but they're playing hard to get. Stay tuned, folks, this AI tango is far from over.

READ THE ARTICLE ON COINTELEGRAPH.

OpenAI unveils ChatGPT plugins, but there's a catch

Well butter my biscuit! OpenAI's letting ChatGPT sprout some extra arms with a bunch of fancy third-party plugins. But don't get all riled up just yet; these new bells and whistles come at a cost. You gotta cough up 20 big ones a month to be a ChatGPT Plus subscriber.

Now, if you're willing to splash the cash, you can get this AI to do all sorts of new tricks. Wanna play trivia, book a trip, or find a bite to eat? No problemo. How about getting stock quotes or finding real estate? You betcha. The sky's the limit with plugins from big shots like Expedia, Zillow, and OpenTable.

OpenAI's out there fighting tooth and nail to stay ahead of the pack, with the likes of Microsoft and Google itching to get a piece of the AI pie. And how do they plan to win? By making the paid plan so darn irresistible, you'd have to be nuttier than a fruitcake to stick with the free option.

Want to give these new toys a whirl? Just head to your chat screen settings, flip the switch for Plugins, and you're in business. You can install as many as you want, but you can only play with three at a time. Then just type a prompt related to your plugin, and Bob's your uncle.

For example, ask about two-bedroom condos in a certain zip code and you'll get listings from Zillow. Or ask to play a trivia game about old Hollywood, and you're in for a swell time. Just remember, no money, no honey - so make sure you're a paid subscriber.

READ THE ARTICLE ON ZDNET.

Amazon is building an AI-powered ‘conversational experience’ for search

Alright, so here's the skinny: Amazon's stepping into the ring with AI chatbots, aiming to soup up its product search. The bigwigs over there reckon it's gonna be as big a game-changer as the Mosaic browser was back in the day. You remember Mosaic, right? No? Well, you're in for a treat, kid.

So, what's the hoopla about? This new-fangled "interactive conversational experience" will answer your questions, compare products, and even toss out personalized suggestions. It's like having your own personal shopper, just without the annoying small talk.

Now, you gotta wonder why Amazon's in such a mad dash. Well, Google just pulled a rabbit out of its hat with an AI-powered search thingamajig that spins up buying guides from a single search. Amazon ain't about to let Google steal their lunch.

But hold your horses, we ain't sure when this new gizmo is gonna roll out, or what it's even gonna look like. We asked Amazon, and all they gave us was some mumbo-jumbo about "investing in generative AI." And honestly, with the current state of AI chatbots, I'm not holding my breath for something mind-blowing.

But hey, it looks like chit-chat shopping is coming to Amazon sooner or later. Get ready for an even more chaotic search experience. Here's hoping they give us an opt-out button, like Google's doing with their fancy AI search.

READ THE ARTICLE ON THE VERGE.

Stability AI unveils new tool for generating animations with just a text prompt

Alright, buckle up, folks. Stability AI, a tech whiz-kid on the block, has shown off a new toy, the Stable Animation SDK. Now, this ain't your grandpappy's sketchpad. This thing can whip up animations from plain text prompts. Yes, you heard it right, just text.

So here's the scoop: You can type in a few words, say, 'fire-breathing dragon' or 'sunset over the Grand Canyon', and this tool will craft an animation for you. You can even tinker with the color, size, and texture of the animations. It's like having a personal Picasso that can decode your brainwaves into art.

But don't just start thinking this tool is as easy as pie. Unlike other image creators, you gotta take a few more technical steps. You can't just twiddle your thumbs and expect a masterpiece.

Stability AI's big cheese, Emad Mostaque, seems over the moon about this. They want to use AI to spark people's creativity, like giving fireworks to a bunch of kids on the Fourth of July. Only time will tell if this new tech will be a hit or if it'll fizzle out faster than a sparkler in a rainstorm.

READ THE ARTICLE ON FOX BUSINESS.

Cloudflare unveils Cloudflare One for AI to enable safe use of generative AI tools

Cloudflare just rolled out its latest "Cloudflare One for AI" toolkit. Designed to keep your business' toys and your customer's secrets safe, it's like a bouncer for your AI playground.

You see, generative AI can be like a toddler with a smartphone. It's enthusiastic, it's learning fast, but it might accidentally share some things that it shouldn't. In fact, 81% of US execs are sweating over cybersecurity issues tied to generative AI, and 78% are fretting over data privacy.

Cloudflare One for AI aims to step in as the responsible adult. It gives businesses visibility of AI tool usage, stops data from wandering off, and helps manage all those new AI tools your team is experimenting with. It's like having a top-notch security system for your AI tech, giving you more control and fewer headaches.

Now, let's talk about some of the shiny new toys in the box. Cloudflare Gateway allows companies to keep track of who's playing with what AI services. It's like a roll call for your AI users, useful for planning and budgeting. It also provides admins with a clear record of API requests and control over which services get their hands on your precious AI training data.

Cloudflare Tunnel is the equivalent of a secure walkie-talkie line to Cloudflare’s network. The data loss prevention (DLP) service is the hall monitor that makes sure no one's spilling secrets they shouldn't.

Cloudflare's betting big that this suite of tools will help businesses embrace the promises of AI while keeping data leaks and security threats at bay. And according to them, this is the first set of AI guardrails out there.

Lastly, they've got a data loss prevention (DLP) service that scans content leaving employee devices for sensitive information. Think of it as the super diligent TSA agent, ensuring nothing sensitive gets through without a thorough pat-down.

Cloudflare's got some plans for the future too. They’re working on scanning AI tools for misuse and misconfiguration. They also plan to offer more integrations with popular AI tools, all part of their mission to help businesses adopt new tech without creating bottlenecks or running afoul of regulations.

So, there you have it, folks. Cloudflare One for AI, the new kid on the block promising to keep your AI tools in line and your data out of the wrong hands. Time will tell if it's the solution businesses have been waiting for.

READ THE ARTICLE ON VENTUREBEAT.

Together raises $20M to build open source generative AI models

Generative AI is all the rage, and startup Together has just scored a cool $20 million to create open source AI models. The idea is to help organizations bring AI into their production apps without breaking the bank. The folks behind Together have some serious street cred, including a PayPal co-founder and a Cloudera founding employee.

Together's plan is to build a cloud platform for running and fine-tuning open source models, offering big-time compute at "dramatically lower" prices than the big boys like Google Cloud, AWS, and Azure. They're aiming for a more transparent and customizable approach to AI, which they believe will be a big draw for regulated enterprises. While the startup's platform hasn't officially launched yet, they're already planning to grow their team and invest their seed money in R&D, infrastructure, and product development. So stay tuned – Together just might shake things up in the AI world.

READ THE ARTICLE ON TECHCRUNCH.

Samsung, Naver to jointly develop generative AI to rival ChatGPT

Well, looky here, folks. Samsung and Naver, two Korean tech behemoths, are teaming up like Batman and Robin to make their own AI platform. They're eyeballing the big league players like ChatGPT and thinking, "Hey, we can do that too!"

Here's how it works: Naver, Korea's go-to for all things online and search-related, gets a bunch of semiconductor data from Samsung. They feed that to their shiny new AI, which Samsung then spruces up even more.

This AI tool, fluent in Korean mind you, will first be used by Samsung's Device Solutions division (that's the one playing around with semiconductors). The tool's supposed to make its debut as soon as October.

But Samsung's not stopping there. After some field tests, they're planning to roll this tool out across the company, even to the divisions handling their smartphones and home appliances.

What's in it for them? Well, Samsung gets a tool to boost productivity and keep their secrets safe from prying eyes. Naver gets a leg up in the global AI market.

Naver is also working on a new AI platform, HyperCLOVA X, that's tailored for Korean services. They say it's a major upgrade from its predecessor and can understand and respond to Korean prompts better than English-based tools. Take that, ChatGPT!

Samsung and Naver are also working on their own AI chips to compete with big shots like Nvidia. This comes after Samsung had a bit of a boo-boo, leaking internal code on ChatGPT to check for errors. Oops! As a result, Samsung, along with some other big companies, are now telling their employees to keep their hands off AI tools like ChatGPT at work.

So, there you have it. The Koreans are getting into the AI game, and they're bringing their A-game. Can't wait to see how this shakes out!

READ THE ARTICLE ON KED GLOBAL.

SAP and Microsoft Collaborate on Joint Generative AI Offerings to Help Customers Address the Talent Gap

Alright, let me break it down for you, nice and easy. SAP and Microsoft are joining forces to tackle one of the business world's biggest headaches - the talent gap. They're using a big, fancy thing called "generative AI" to help businesses attract, keep, and develop their workforce.

SAP's SuccessFactors is going to buddy up with Microsoft 365's Copilot and Viva Learning. Also, they're bringing in the big guns with Azure's OpenAI Service, known for its smarts in analyzing and creating natural language. This is all aimed at making it easier for companies to hire the right people and keep them around.

It's like bringing a supercomputer to a spelling bee – the integration will take over the humdrum stuff, like job descriptions and interview questions, and even help suggest personalized learning for employees. All in all, it's about making the process smoother, smarter, and less of a pain in the you-know-what.

But don't worry, this AI isn't going rogue. SAP assures us it's all above board, transparent, and respectful of privacy. So, if it works out, it could be a game-changer for tackling the talent gap. So, keep your eyes peeled for what comes next in this high-tech romance. And as always, remember: with great AI power comes great responsibility.

READ THE ARTICLE ON SAP.

The man behind ChatGPT is about to have his moment on Capitol Hill

Sam Altman, the CEO of OpenAI and the mastermind behind ChatGPT, is set to testify before Congress on the risks of artificial intelligence (AI). Altman's appearance comes as concerns grow about job displacement, biases, and misinformation caused by AI tools. He has been praised for his brilliance and forward-thinking, but also faces calls for more caution, including from Elon Musk.

Altman's influence extends beyond AI. He achieved success with the app Loopt, sold for $43 million, and served as president of Y Combinator, connecting with influential figures in the tech industry. He advocates for lower living costs, clean energy, and universal basic income, believing AI-generated wealth could facilitate it.

OpenAI, co-founded by Altman and Musk, aims to ensure safe and beneficial AI development. ChatGPT's success has made Altman a central figure in AI's growth. However, his cautious approach reflects the potential dangers of technology.

As the future of AI approaches, Altman remains uncertain about the outcome. He prepares for worst-case scenarios, including AI attacks. Despite uncertainties, Altman is dedicated to shaping a positive future through responsible AI development.

In summary, Sam Altman, CEO of OpenAI and the brains behind ChatGPT, will testify before Congress on AI risks. Altman's influence, caution, and dedication to a positive future make him a significant figure in AI's growth.

READ THE ARTICLE ON CNN.