YouTube Introduces AI-Powered Dubbing!

Unlock the Power of Your Crypto

Arch is a revolutionary new platform that empowers crypto investors to obtain a single loan collateralized by multiple crypto holdings, all securely held by the leading qualified custodian, BitGo. At Arch, we adhere to strict regulatory frameworks and never touch customer funds, so you can trust that your investments are safe and secure. Experience instant liquidity now with a fixed-rate term loans at 9% APR against all crypto portfolio. Join the Arch community today and take your crypto holdings to the next level.

Learn More

Today:

YouTube is getting AI-powered dubbing

YouTube is looking to add a little AI seasoning to its video potluck. They've got their eyes on Aloud, an AI-based service that gives videos the gift of gab in multiple languages.

So, how does it work? It's simple. Aloud first listens to your video, jots down what's said, and lets you polish up the transcript. Then, it switches on the translator, and voila, your video speaks a new language. And this isn't just some crazy idea on a whiteboard, they're already trying it out with a bunch of video-makers.

Right now, Aloud only speaks a few languages - English, Spanish, and Portuguese. But they're promising more will join the party soon.

YouTube's not stopping at simple translations. They're aiming to have these AI-powered voice tracks match the creator's voice, tone, and even lip movements by 2024. So, keep your peepers peeled for that.

READ THE ARTICLE ON THE VERGE

AWS launches $100M program to fund generative AI initiatives

Amazon is launching a new $100M initiative, the AWS Generative AI Innovation Center. The goal? Pair their in-house brains with customers and partners to fast-track some serious enterprise innovation using generative AI. You might ask, "What's that mean, exactly?" Well, according to Sri Elaprolu, the big cheese at the Center, it's about helping customers dream up, design and launch new AI goodies.

This new initiative aims to guide clients in cooking up their own AI plans, pinpoint high-priority use cases tied to business value, and lay the groundwork to kickstart those plans. It's a sort of 'you-scratch-my-back-I'll-teach-you-AI' deal. Clients will get their hands on free workshops, training, and access to AWS products like CodeWhisperer and Bedrock. And who are the lucky customers? Well, those who've already been tapping on AWS' door for AI assistance get first dibs, especially if they're from sectors like finance, healthcare, media, and a few more.

Despite a somewhat shaky start with their Bedrock platform, AWS has managed to rope in some big fish in the AI pond, like Stability AI and AI21 Labs. They're also going head-to-head with Google Cloud and Microsoft in the scramble for AI supremacy. The stakes? A market that could be worth a whopping $110 billion by 2030.

READ THE ARTICLE ON TECHCRUNCH

ClickUp Introduce AI-powered Assistant Tailored To Your Work

ClickUp is dipping its toes into the AI pool and, by gum, it's one heck of a cannonball! Just in time for World Productivity Day, it's introducing ClickUp AI, the first AI assistant made to make work, well, less work.

The new AI has over 100 built-in tools, ready to do the nitty-gritty for you. It's got everything from creating templates to automating workflows. Want a quick-and-dirty summary of your docs and tasks? It's on it. Need action items pulled from your meetings or comments? Consider it done. Don't feel like reading long comment threads? It'll serve you up the juicy bits.

ClickUp AI will whip up a daily or weekly standup report based on your activity. And for the wordsmiths among us, it'll help overcome writer's block and draft your emails, blog posts, and customer responses.

Staying on top of your projects is also easier. Just pick the tasks, pick the time period, and voila, you've got a shiny progress report. It can even use past tasks to automatically create subtasks for you, all by its lonesome. You can access ClickUp AI anywhere within the platform, whether it's a task or a doc.

READ THE ARTICLE ON CLICKUP

Microsoft Unveils Predictive Targeting, AI-Based Advertising Tool

Microsoft has pulled the curtain back on its new tool, Predictive Targeting. It's like a digital crystal ball that uses artificial intelligence (AI) to help advertisers find folks who are most likely to click on their ads.

This tool doesn't just look at the advertisers' ads and web pages, it also takes a peek at Microsoft's massive pile of user data to spot potential new audiences. Microsoft claims it can boost the chance of people actually buying stuff by an average of 46%.

Predictive Targeting offers a promising way to reach the right audience and save advertisers some serious clock. But don't put all your eggs in one basket. Other targeting tools may also need to come to the party to avoid unexpected guests. And yes, you can turn this tool off if you feel the need to call the shots.

READ THE ARTICLE ON SEARCH ENGINE JOURNAL

Inflection debuts its own foundation AI model to rival Google and OpenAI LLMs

AI startup Inflection, just unveiled its Pi conversational agent, fueled by a large language model, named Inflection-1. Similar in size to OpenAI's GPT-3.5, this new tool is making claims that it's as good, or even better, than others in its class.

This Inflection-1 is quite the show-off in middle and high school level tests and common-sense questions. It's a little less savvy when it comes to coding, where GPT-3.5 and especially GPT-4 have the upper hand.

Inflection hinted at a bigger model, something on par with GPT-4, cooking in the back room, but they're keeping it under wraps until it's ready to shine.

READ THE ARTICLE ON TECHCRUNCH

Armilla AI debuts AutoAlign, allowing enterprises to fine-tune their AI models to block hallucinations, harm

Armilla AI, a three-year-old outfit backed by the high-profile YCombinator, is stepping in with a new product called AutoAlign. It helps adjust AI models to prevent them from making stuff up or giving harmful responses.

AutoAlign can be used even by folks with little tech savvy. It’s a tool for organizations to fine-tune their AI responses and remove biases. Another trick up AutoAlign's sleeve is setting up 'guardrails' for AI models. These are basically invisible fences that keep AI from spilling secrets or making harmful suggestions. For example, an AI could be fooled into explaining how to make napalm, a dangerous firebomb. AutoAlign, however, would spot the foul play and put a stop to it.

These guardrails can also stop the AI from spinning yarns or hallucinations, like inventing non-existent court cases or facts. And when it comes to sensitive info, AutoAlign keeps it on lockdown, preventing any leaks. AutoAlign is gearing up to help make AI safe and responsible because as it stands, there are more people building AI than there are making sure it's safe.

READ THE ARTICLE ON VENTUREBEAT