Google and Character.AI Join Forces

Today:

Google in talks for a major investment in AI start-up Character.AI

Google is chatting up Character.AI about throwing a ton of cash their way, like hundreds of millions. Character.AI, started by a couple of ex-Google folks, Noam Shazeer and Daniel De Freitas, is all about letting folks chat with fake celebs or whip up their own chatbots. They're free to use, but if you're itching to cut the line, it's $9.99 a month.

Most of Character.AI's users are young guns, between 18 and 24, making up about 60% of the crowd hitting up their site. They're trying to stand out in the crowd with fun AI pals, different from other big names like ChatGPT and Google's Bard.

Meanwhile, Character.AI is also chit-chatting with venture capitalists, aiming for a sweet $5 billion valuation. Just this March, they bagged $150 million with a value tag of a billion bucks.

Google and Character.AI have been tight for a while, with Character.AI using Google's cloud and fancy tech for training their AIs. But, both sides are zipped up about the deal talks, which could still change.

READ THE ARTICLE ON REUTERS

GitLab expands its AI lineup with Duo Chat

GitLab, earlier this year, rolled out Duo - a cool set of AI tools to help coders do their thing better and faster, like summing up issues, making sense of big projects, and giving a hand with coding and security stuff. Now, they're bringing Duo Chat to the party. Think of it as a ChatGPT-like chatbot that lets developers chit-chat with Duo to use all its nifty features in a more hands-on way. It's still in the testing phase, though.

GitLab's now using Anthropic's Claude for their chat backend, not Google, which they've been buddies with for other AI features. DeSanto explained they needed something with a bigger brain to handle chat's heavy lifting and keep up with longer conversations.

GitLab's also shaking hands with Oracle for some heavy-duty cloud computing power. They're not picky about their cloud partners - it's like a tech love triangle with Google, Oracle, and Anthropic all in the mix.

Duo Chat's already got some cool tricks up its sleeve, like explaining code and making it better. More features are coming down the pipe, like creating test cases and breaking down security issues.

READ THE ARTICLE ON TECHCRUNCH

Adobe sells fake AI-generated Israel-Hamas war images – then the news ran them as real

Adobe's selling these wild AI-made pictures that look like they're straight out of the Israel-Hamas conflict, but they're totally not real. Last year, Adobe was like, "Hey, we'll sell your AI-made snaps and give you a cut." So, folks can make anywhere from 33 cents to over 26 bucks each time one of these faux war photos gets downloaded.

Here's the kicker: These AI tools are so good now, anyone can whip up some realistic-looking images. Some people got the bright idea to make fake war pics and sell them on Adobe Stock. Even though Adobe labels these as AI-generated, that tag sometimes gets lost in the shuffle. Smaller news outlets have been using these pics in their stories, not always saying they're made-up.

Adobe and some other big names are working on this thing called Content Credentials. It's supposed to show where an image came from, but it's still not out there in the real world yet. It needs everyone - social networks, publishers, artists, and techies - to play ball for it to work right.

READ THE ARTICLE ON THE REGISTER

Picsart launches a suite of AI-powered tools that let you generate videos, backgrounds, GIFs and more

Picsart, already big in the photo-editing game and backed by SoftBank, is rolling out something called Picsart Ignite. This thing's packed with 20 AI-driven tools. We're talking stuff for whipping up ads, social media posts, logos, you name it. It's got everything from video generators to GIF makers, plus fancy features like "AI Expand" for adding more detail to pictures, and "AI Object Remove in Video" to get rid of unwanted stuff in your clips.

There's also this cool "AI Style Transfer" to give your visuals an artsy touch and an "AI Avatar" tool for creating lifelike avatars. If you're feeling quirky, you can even turn yourself and your pet into superheroes or royalty. Picsart isn't new to this AI game – they've already got tools like AI Writer and AI Image Generator, cranking out over 2 million images daily.

Picsart Ignite is now up for grabs on the web, iOS, and Android. This launch is part of a bigger trend where all sorts of platforms are jumping on the AI bandwagon. Earlier, Canva added some AI tools, and Figma's got new AI features for its FigJam whiteboard. Picsart’s been on this AI ride for a while – they even dropped an app called "SketchAI" that turns photos and drawings digital, with a range of artistic styles to play with.

READ THE ARTICLE ON TECHCRUNCH

Health AI startup Cercle debuts with backing from Sheryl Sandberg

Sheryl Sandberg, the big shot who used to call the shots at Facebook, is throwing her weight and wallet behind a new health tech startup called Cercle. This ain't your average tech gig; Cercle's playing a whole new game, using AI to shake up women's health, especially in the fertility department.

Cercle's got this fancy platform that takes a jumble of medical records and test results and makes sense of them for fertility docs and researchers. They're hoping this smart tech can help doctors whip up tailor-made treatments for their patients and speed up the hunt for new meds.

Say you're looking to freeze your eggs. Cercle's system lets your doctor peek at info from other folks who've done the same. It's all anonymous, but it gives docs a heads-up on what's worked before, helping them come up with a plan that's just right for you.

Cercle's buddying up with Eurofins Genoma, a big name in genetics, and US Fertility, the largest fertility network around. For now, they're zeroing in on fertility, but they've got plans to branch out into more women's health areas down the line.

READ THE ARTICLE ON CNBC

Ghost, now OpenAI-backed, claims LLMs will overcome self-driving setbacks — but experts are skeptical

Ghost Autonomy, this new kid on the block, is betting big on these fancy AI models called "multimodal large language models" (LLMs) to fix the troubles with self-driving cars. They've teamed up with OpenAI and bagged a cool $5 million to play around with this tech.

Ghost, though, is all gung-ho. They say their software can make sense of tricky driving scenes, like those messy construction zones, using LLMs. They're testing this stuff, hoping to prove that these AI brains can handle the wild world of driving.

Meanwhile, the experts are shaking their heads. They're pointing out that even basic tasks trip up these AI models, let alone something as dicey as driving. No one's found a magic bullet in computer science yet, they say, and slapping LLMs onto self-driving might be jumping the gun.

OpenAI's bigwigs and Ghost's head honcho ain't backing down, though. They're jazzed about the possibilities, thinking these AI models might just crack the code on understanding complex driving scenarios. But as Ghost admits, there's a long road ahead before this tech is ready for prime time.

READ THE ARTICLE ON TECHCRUNCH

Ozone raises $7.1M to scale its AI-powered collaborative video editor in the cloud

Ozone, a fresh startup, just bagged $7.1 million to grow its AI video editing tool in the cloud. Think of it as a sidekick for content creators, making tedious editing tasks super quick. Max von Wolff, the brain behind it, wants to beef up creativity, not replace it. He got the idea while studying at MIT, ditched school, and jumped into Y Combinator to make it happen.

This tool is pretty slick. It's got AI that suggests edits, tons of video editing tricks, endless cloud storage, and lets folks work on videos together in real-time. You can do stuff like cut out awkward silences, slap on captions, tweak colors, and throw in tunes with AI's help.

Ozone’s team is a mix of smart cookies from big names like Adobe and Disney. The tool's a godsend for folks making quick videos for platforms like TikTok or YouTube. And they're planning to add more bells and whistles for longer stuff like documentaries.

READ THE ARTICLE ON TECHCRUNCH

PwC’s $1B Investment Will Give Every Worker AI Training—And Staffers Access To Chatbot Assistants

PwC, a big-time accounting and consulting firm, is throwing a whopping $1 billion into AI over the next three years. This move came after their top tech guy, Joe Atkinson, got pumped about AI at a big meeting in Switzerland. They're not just splurging on tech; they're also training their 75,000 employees on AI, partnering with Microsoft and OpenAI, and even got plans for their own AI-driven tools.

This push isn't just a PwC thing. Other heavy-hitters like Deloitte, EY, and KPMG are also betting big on AI. They all see AI as a game-changer, especially for jobs that need a lot of brainpower like legal, audit, and consulting work. Atkinson even says AI is the first thing in 30 years that could really speed up and improve how knowledge work gets done.

Here's how PwC is rolling with it: after the big meeting in Switzerland, they decided to dive headfirst into AI instead of sitting back and watching. They started training their people on the ethics and smart use of AI, with 60% already on board. They're using AI internally with a tool called “ChatPwC” and trying it out in all sorts of ways, like helping tax folks with asset values and letting business analysts quickly sum up long reports.

They're even testing AI with their U.K. team for tax, legal, and HR stuff, thanks to a startup called Harvey. This could go global next year, covering 50 places. Despite the industry slowing down, PwC's chief people officer, Seals-Coffield, says they're not expecting AI to cut jobs. They see AI more as a helper for their human workers, not a replacement.

READ THE ARTICLE ON FORBES

Why Are We So Afraid of an AI Apocalypse?

So this article is breaking down why folks are jittery about AI turning into a doomsday machine. It's like, back in the day, we were scared of saber-tooth tigers and stuff, and now it's rogue AI in movies freaking people out. The thing is, AI's just a tool – it's not gonna grow a mind of its own and go berserk. What's really sketchy is how the big tech honchos might be hyping up these scare stories. They're playing this angle where they're like, "Only we can handle this beast," to keep the little guys out and have the big slice of the pie.

Then there's this guy, Andrew Ng, calling out these tech giants. He thinks they're making a mountain out of a molehill about AI being the end of the world, just to get the government to play ball their way and keep their top-dog status.

The real baddies in this story? Laziness and fake news. If we just swallow up these tall tales without thinking and let lies run wild, we're setting ourselves up for trouble. The article's saying we gotta be sharp, not let fear run the show, and make sure AI's working for everyone, not just a few big shots.

READ THE ARTICLE ON PSYCHOLOGY TODAY