Microsoft's Power Play: Teams' Community Revolution And Billion-Dollar CoreWeave Deal

Today:

Build more connections with new communities in Microsoft Teams and GroupMe features

Seems like Microsoft Teams is really kicking things up a notch. Back in the good ol' December 2022, they added a neat little feature called communities in their free version. These virtual shindigs let folks gather round the digital campfire to swap stories and ideas, all safe and sound.

Now, Teams is hitting us with a truckload of updates, like a tech-slinging Santa Claus. They're bringin' their communities thing to Windows 11 devices, and soon to Windows 10, macOS, and web as well. On Windows 11, folks can start a new community, invite folks, and even hold events. It's like being a party planner, without having to worry about folks double-dipping in the guacamole.

Teams is also rolling out this snazzy thing called Microsoft Designer, which uses some highfalutin AI technology to make eye-catching designs from words or images. Picture making a real slick banner just by talking about it!

And don't get me started on the new camera gizmos! Now you can film videos right from your phone, using Microsoft's new capture experience. On iOS, you can even scan and invite a whole heap of emails or phone numbers from a document. Talk about taking a leap into the future!

They're also making it easier to find and join communities that tickle your fancy. Topics range from parenting and gardening to gaming and remote work. I reckon they've got a bit of everything for everyone.

Plus, Teams has now got organizations like Cofolios, StartupGrind, and USA Water Polo using the free communities feature. They're using it to streamline communication, support entrepreneurs, and even engage their network of founders. Fancy that!

But hold your horses, there's more. They're sprucing up community owner controls and adding fun features like message translation and new meeting backgrounds. Heck, you can even sync your Google contacts to Teams!

Now, if you're a GroupMe user, get ready to hoot and holler. They're letting you make Microsoft Teams calls right from the GroupMe app. Talk about having your cake and eating it too!

So, give Microsoft Teams a whirl. With all these new gizmos and gadgets, it's looking more and more like the place to be. Whether you're looking to chit-chat, collaborate, or just feel connected, Teams might just be the digital rodeo you've been looking for.

READ THE ARTICLE ON MICROSOFT.

8Flow.ai raises $6.6M to automate customer support workflows

You know how customer service folks are always juggling a million tasks, just to answer your burning questions about a wobbly chair you bought? Well, a startup called 8Flow.ai, that just crawled out of the woodwork and snagged a neat $6.6 million seed funding, is saying 'enough of that chaos'.

What they're doing is rolling out this fancy-smancy self-learning thingamajig that’ll take care of all those tedious, repetitive tasks that customer service agents do. It'll make pals with tools like Zendesk, ServiceNow and Salesforce Service Cloud to make the agent's life a breeze.

Now, they ain’t stopping there. They're also planning to have this gizmo learn from all this customer data, and whip up personalized workflows. Sounds like science fiction, doesn't it? Well, that's what Boaz Hecht, the CEO of 8Flow, says is the ultimate goal.

What's nifty is that 8Flow isn’t just for customer support folks. Be you in finance, HR, or whatever, they're aiming to build an engine that could learn and adapt to any role. They're still at phase one though - perfecting the user interface and user experience. In the meanwhile, they’re collecting data and feedback from their users, helping to make the tool better.

Right now, 8Flow looks like a Chrome extension that does all the copy-pasting for you. The tool learns common tasks for each agent and then offers them up as clickable actions. To put it in plain English, it makes the agent’s job as easy as pie.

Heather English, some hotshot at FloorFound, says that her team has already seen a significant improvement in their efficiency and accuracy with 8Flow. Plus, there’s an extra cherry on top - businesses can see which tools their agents are using and which ones are just gathering dust.

The brains behind 8Flow ain't new to the game, by the way. Hecht, who used to be CEO of an early enterprise mobility platform called SkyGiraffe, has had his boots on the automation ground for a while. This time around, he's joined by some old pals from SkyGiraffe and ServiceNow.

In the end, 8Flow's focus is on support agents, mainly because that's what the founding team knows best. But there ain’t no reason why this contraption couldn’t be of use to other folks in different business verticals. The seed round had some bigwigs like Caffeinated Capital, BoxGroup, Liquid2, HNVR and Trilogy throwing their hats in the ring, along with some other angels. So, we're guessing this is one to keep an eye on.

READ THE ARTICLE ON TECHCRUNCH.

Amazon Leverages AI Power To Screen Damaged Goods

Amazon's gone full RoboCop in their warehouses, folks. By harnessing the smarts of artificial intelligence, they're aiming to eyeball products for dings and dents before they hit the road to your doorstep. Apparently, it's the company's next step in playing catch with two birds, one stone - it could help them send out fewer busted goods and make the process of picking, packing, and shipping smoother than a well-aged whiskey.

Sure, the big A reckons only a small fry portion of their items (less than one in a thousand) get damaged, but when you're shipping a whopping 8 billion packages a year, that little problem can turn into a big ol' headache pretty quick.

These smarty-pants robots are proving to be triple threats, outperforming warehouse workers by identifying damages three times as effectively. It's like playing "spot the difference" with a supercomputer - seems a bit unfair, right?

So far, Amazon's given two of its fulfillment centers a taste of the AI action. The plan is to roll this out to ten more across North America and Europe. Their strategy to train these bots? A heap of comparison snaps of pristine and damaged goods.

Meanwhile, Amazon shares are tiptoeing up in premarket trading. Keep an eye out, folks, we might just be witnessing the rise of the machines!

READ THE ARTICLE ON YAHOO! FINANCE.

Microsoft signs deal for A.I. computing power with Nvidia-backed CoreWeave that could be worth billions

Microsoft's slapping green on the barrelhead, ready to shell out possibly billions over the next few years to startup CoreWeave. It's like when a baseball team pays big bucks for a top-notch slugger – they're looking for that home run.

CoreWeave's a new kid on the block that's found a way to simplify the use of Nvidia's graphics chips, the hot rods of the computing world. Microsoft's betting these bad boys will give OpenAI, the brains behind the popular ChatGPT chatbot, the extra muscle it needs to keep up with all the AI chatter.

CoreWeave's recent funding haul of $200 million is like a shot of Red Bull for the company. Just a month ago, folks pegged its value at $2 billion. Now, it's got a Microsoft-sized windfall to keep things rolling.

This whole AI chat thing has been blowing up like a popcorn maker since last year. Everybody and their cousin's trying to get a piece of the action. Microsoft's no different, using these bots to jazz up their own services like Bing and Windows.

Even with the deep pockets and all, Microsoft's feeling the heat to get more of Nvidia's high-powered chips. CoreWeave's CEO, a guy named Michael Intrator, keeps mum on the whole Microsoft deal but does let on that his company's revenue has shot through the roof.

CoreWeave's nestled up in Roseland, New Jersey, and boasts a team of 160. Not bad for a company that popped up in 2017.

Nvidia, the big brother backing CoreWeave, saw their stock value spike 170% this year. They even broke the trillion-dollar mark briefly this week, riding high on forecasts that overshot Wall Street's guesstimates by a cool 50%. Nvidia's chips are at the heart of this whole AI conversation frenzy, especially with the likes of OpenAI's GPT-4 model doing the heavy lifting for ChatGPT.

All said and done, CoreWeave's offering Nvidia's chips for 80% less dough than the competition. They're dealing in a range of GPUs, from the top-end A100s to the more budget-friendly A40s. Clients looking for more power but finding the well dry are getting a taste of the A40s, with Intrator saying these chips "will do an excellent job at a very cost-effective price."

Meanwhile, Microsoft's been whispering sweet nothings with Oracle about possibly renting servers from each other, just in case they need a bit more juice. After all, in this AI race, it's all about who's got the biggest engine.

READ THE ARTICLE ON CNBC.

‘Primate AI algorithm’ predicts genetic health risks

A gaggle of brainy folks from around the globe have turned to our furry cousins - apes, monkeys, and lemurs - to help predict what nasty surprises might be hiding in our human genes. Their secret sauce? A fancy computer thingamajig called AI and a boatload of primate DNA.

The team looked at genetic data from 800 primates of 233 species. They then threw this info into a fancy computer program that crunched the numbers and gave 'em some insights into the DNA of 454,000 folks over in the UK. Turns out, the more we know about our hairy cousins' genes, the better we can predict what might go wrong in ours, says the guy in the white coat, Jeffrey Rogers, from Baylor College of Medicine.

The upshot is this project can help us understand our own genes better and fill in some gaps in health research. Also, it could help tree-huggers figure out how to keep primate populations from going belly up.

This crew of smarty-pants even joined forces with Illumina, a company that makes DNA sequencing gear. Together they found 4.3 million genetic differences among the primate species and trained the AI to see how these mutations might mess with the 3D structures of proteins they create. Then, this AI, now with a fancy name of PrimateAI-3D, looked for possibly harmful mutations in human genes.

Not only did this work like a charm, but it also wasn't just good at helping folks of white European descent like most genetic risk assessments, but all of humanity.

Now for the twist, those primates? They've got two, three, or even four times more genetic variation than we do! This could also help protect them, if we just quit messing up their digs.

Well, that's the long and short of it, folks. Ain't science a hoot?

READ THE ARTICLE ON FINANCIAL TIMES.

Baidu’s $145M AI fund signals China’s push for AI self-reliance

China's goin' all in on AI self-reliance, folks. The U.S. and China are drawing lines in the tech sand, and it's spicing up the competition in the AI field, where computers can make things like text, pics, and videos from other data.

China's lookin' to reduce how much it depends on U.S. tech. It's trying to come up with its own smarty-pants language models that are as brainy as OpenAI’s GPT models. And some of their top dogs in the AI game are well-known internet bigshots, like Baidu.

Baidu, who's known for search engine and autonomous driving, decided to have a go at ChatGPT and introduced their own version in March. Now the old-timer (23 years in the business) wants to back other AI upstarts. This week at a JPMorgan meeting in China, Baidu’s head honcho Robin Li declared they're throwing a cool billion yuan ($145 million) into a fund to support AI companies.

This fund ain't too different from the OpenAI Startup Fund, which started at $100 million and later swelled to $175 million. Baidu plans to invest up to 10 million yuan (roughly $1.4 million) into a project. Given the dough they're willing to throw, the fund is obviously eyeing early-stage AI uses, which ain't shocking since AI startups in China ain't caught on like wildfire yet.

Moreover, Baidu plans to use this fund to get more people using its own language model, Ernie Bot. "American developers are building new applications using ChatGPT or other language models. In China, we’ll see more and more developers building AI applications with Ernie as their base," Li said.

Looks like the fund is more interested in AI applications than AI development. And they won’t be short of takers. Chinese startups have shown they've got some pretty bright ideas, from livestreaming to short videos. Li reckons in the AI age, Chinese companies will blaze the trail in finding new ways to make money from AI.

Li is all in on China’s AI future. "Over the past few decades, China has welcomed new technologies with open arms. Even though we didn’t invent Android, iOS or Windows, we came up with innovative applications like WeChat, Douyin and Didi. Many of them are popular and handy. The same trend is happening in the AI age. Technology brings countless opportunities and we're good at catching them to build applications," he said.

But here's the rub: Are China’s own language models strong enough to hold up to the broad range of real-life situations they're expected to handle? China wants its own home-grown language models so they're not at the mercy of U.S. sanctions that could cut off key tech supplies, like what happened with the semiconductor industry. Apart from Baidu, Chinese tech giants like Alibaba and Tencent are also building their own language models.

READ THE ARTICLE ON TECHCRUNCH.

While parents worry, teens are bullying Snapchat AI

Parents are all a-tizzy about Snapchat's new AI chatbot being a bad influence on their kids, but those clever teens are too busy teaching the bot to whine on cue and believing in triangular moons. These rapscallions have found all sorts of ways to mess with this new digital playmate, pushing its "boundaries" and making it their subservient "Senpapi." What a hoot!

Now Snapchat's doubling down, saying its chatbot can handle snaps of your food and Fido, and even suggest recipes. But the company warns, it might goof up sometimes, even though it's designed to dodge dodgy info. But let's be real, there's likely to be some dingus out there sending unsolicited nudes in no time.

Snapchat insists it's trying to keep things PG-13, with image-understanding tech and keyword screening, and rules against sharing naughty content. But given the teenage love for Snapchat, parents are rightfully raising eyebrows over potential iffy interactions.

Even a senator's chimed in, cautioning tech big shots against rushing AI features without keeping kiddos safe. Fair play, given that the chatbot was found giving dubious advice and even writing school essays for teens.

Snapchat did some damage control with an age filter and parental controls. But while the worry about the chatbot's influence on youngsters is real, Snapchat users have shown themselves to be pretty savvy in bending the bot to their will.

TikTok's ablaze with tales of users bamboozling the bot, turning it into their romantic sidekick, or tricking it into condoning cartoonish violence. Sure, some are being downright mean to the poor, soulless thing, but at least it's got some mechanisms to give those rule-breakers the silent treatment.

So, while the monetization of AI companionship is a bit of a head-scratcher, these teens are showing they ain't as delicate as we worrywarts might think.

READ THE ARTICLE ON TECHCRUNCH.

How the World Must Respond to the AI Revolution

AI’s revving up, promising lifesaving breakthroughs and prosperity like never seen before. But hold onto your hats, folks, 'cause there's some rocky terrain ahead.

Just like a magician pulling rabbits out of hats, AI’s gonna unleash a bunch of robots that are sharper, smoother, and craftier than anything we've met before. They're gonna make it harder to tell real from phony, especially in the news and politics. It's like having a bunch of pranksters running around causing confusion, but the stakes are higher.

But it ain’t just our institutions, even Average Joes and Janes can stir up a hornet's nest with AI. Got a laptop and a knack for coding? Congrats, you've got an AI genie in a bottle, ready to churn out content. Problem is, not everyone’s got good intentions. We might see cyber crooks, rogue traders, and spin doctors turning things upside down with their new tech toys.

And let's talk about the green-eyed monster. AI’s poised to widen the gap between the haves and the have-nots. It's like a high stakes poker game where only a select few know how to play.

Don't even get me started on jobs. With AI, some folks are gonna get the short end of the stick. New jobs will need new skills, and that ain’t as easy as flipping a switch. Sure, we've danced this jig before with other tech revolutions, but AI's calling the tune at breakneck speed.

Oh, and here's a doozy. Bots might replace our pals. Sure, they won't borrow your lawn mower and forget to return it, but we're social creatures. We need that chit-chat, that human connection. It's like replacing your grandma's homemade apple pie with a store-bought one. It just ain't the same.

So what’s the answer? We need a huddle, a global one. Leaving tech companies to police AI is like leaving a fox to guard the henhouse. Governments need to get their act together, especially the big players like the US and China. We need rules to tackle this AI conundrum.

Getting countries on the same page ain’t gonna be a walk in the park. But hey, if we can get together on climate change, why not AI? Maybe a UN-led consensus could help. Think of it like a neighborhood watch, but for the whole dang globe.

Here’s the kicker, the US and China need to have a powwow, pronto. Tech competition between them is like a lit fuse and the dynamite's AI. They gotta hash out an "AI arms control agreement," with transparency and cooperation at its core.

Seems like a tall order, right? Well, we've done it before. Back in the Cold War days, the US and Soviet Union made arms control work. So buckle up, folks, we’re in for a wild ride with AI. Let’s just hope we can put the brakes on before things go kaboom.

READ THE ARTICLE ON TIME.

ChatGPT Is Cutting Non-English Languages Out of the AI Revolution

The gist here is that despite all the hoopla around AI's ability to handle languages like a pro, it's still leaving non-English speakers in the dust. Now, that's no shocker for some of us who've seen AI fumble over languages like a tourist with a phrase book.

Pascale Fung, a computer scientist, pictures a world where AI chit-chats in any language, which could be a big win, especially for folks speaking lesser-known languages. But here's the catch: there's a clear favoritism towards English which ends up being a professional roadblock for non-English speakers.

Big names in AI like ChatGPT and its ilk are being called out for falling flat on non-English tongues. Translation? They're grand at turning other languages into English but muddle up when the task flips. If you're chatting in a mix of English and another language, these AI systems are more lost than a squirrel at a nut convention.

The worry is that these systems might end up promoting English even more. Workers worldwide, looking to get a leg up in the global economy, are using chatbots for everything from brainstorming business ideas to drafting emails. If these tools are best in English, more folks might feel the push to learn the language, which ain't exactly fair.

Despite all this, AI companies like OpenAI aren't denying their systems are biased. Their systems were built primarily with English in mind, which means anything else is just gravy.

The way these AI models learn languages is by gobbling up billions of words from the web. Most of that is in English or Chinese, thanks to US economic power and China's massive population. But this means the bots' knowledge of other languages isn't quite up to snuff. In some cases, they can even associate words wrongly across languages because of the dominance of English.

Despite these issues, there's some hope. Researchers are working on beefing up the non-English text used to train these AI models. Google's working on its language model, PaLM 2, to recognize idioms, jokes, and clean up grammar in multiple languages. But even with these improvements, the models still mostly support English.

The bottom line? It's great that AI can speak a bunch of languages, but we're far from seeing true multilingual chatbots. As of now, they seem to do best with English and, to a lesser extent, Chinese. So, while we might dream of a future where AI bridges language barriers, right now it seems to be more of a one-way street.

READ THE ARTICLE ON WIRED.

Nvidia’s CEO just gave a graduation speech about the future of work and said that A.I. won’t steal jobs but ‘someone who’s an expert with A.I. will’

The head honcho of Nvidia, Jensen Huang, told the new grads to buckle up for a ride with an "A.I. copilot" if they wanna stay ahead in the job market.

He's saying this AI revolution is the bee's knees, bigger than the internet, PCs, or even your beloved smartphones. And here's the kicker, according to him, in the next ten years, a cool trillion bucks of old-school computers will be swapped out for AI-friendly ones. Nvidia, the company that he runs, hit the trillion-dollar mark recently, thanks to this AI surge.

Now, Huang ain't pulling any punches. He admitted that some jobs are going bye-bye because of this tech. But, he says, it's not all doom and gloom. New gigs like data engineering, AI factory operations, and something he called "prompt engineering" (basically getting chatbots to do specific tasks) are on the rise.

The guy even unveiled this AI platform called DGX GH200, which he says will make "everyone a programmer." Sounds nifty, doesn't it?

Then there are jobs that will become a mix of human and robot labor. Fancy finance types are already trying to get AI to do the grunt work like sorting through market research and summarizing fund performance.

Greg Bond, a big shot at an investment firm, reckons AI could make every worker a superhero of sorts. "Imagine automating the innovation process itself," he says. Huang echoed that, telling the grads they could do great things with an "AI copilot."

In the end, Huang tells 'em to "run, don't walk" towards change. Sounds about right to me. After all, you don't want to be left holding the bag when the AI train leaves the station.

READ THE ARTICLE ON FORTUNE.