Google's Copyright Protection

Today:

Google promises to take the legal heat in users’ AI copyright lawsuits

Google's throwing its legal muscle behind users of certain AI products, vowing to shield them from copyright lawsuits. The tech giant has highlighted seven AI products, like Duet AI and Vertex AI features, saying, "We got your back" if copyright sharks come biting due to the AI using or generating protected content.

They're rolling with a two-step plan to absorb legal troubles tied to their training data and AI-generated results, while also making it crystal clear: intentional mischief isn’t covered. Other big tech names, Microsoft and Adobe, have also stepped up, saying they’ll fend off legal woes for their AI users. This all comes amidst a flurry of lawsuits, even from big-name authors, challenging companies on AI and copyright grounds.

READ THE ARTICLE ON THE VERGE

Lakera launches to protect large language models from malicious prompts

Swiss startup Lakera, armed with a crafty interactive game named Gandalf and a fresh $10 million funding, has stepped into the spotlight, aiming to shield large language models (LLMs) and enterprises from cyber shenanigans like "prompt injections" and data leaks. Gandalf, while on its face a whimsical game, allows users to try to trick an LLM into spilling a secret password, feeding crucial data about its vulnerabilities back to Lakera.

This isn’t just for kicks, it’s a savvy move to bolster Lakera Guard, their main product, aiming to provide an extra layer of security against both naughty actors trying to milk LLMs for unauthorized access, and potential data spillage that could be harmful or a privacy no-no. Lakera, not just sticking to security, is also diving into the tricky waters of content moderation and data privacy, while also keeping a keen eye on emerging AI regulations in the EU.

READ THE ARTICLE ON TECHCRUNCH

Edge AI: The AI No One Is Talking About

So, imagine all that computer brainpower but it’s happening right where the action is - that’s edge AI. It’s like having a super-smart assistant that doesn’t need to run back to the main computer (cloud) every time it needs to make a decision.

Imagine self-driving cars making zippy, safe choices without taking a data detour through the cloud. Quick, on-the-spot, and efficient. It’s making business tools, like your vid calls on Zoom or Teams, run smooth and secure by handling stuff right there on your device.

READ THE ARTICLE ON FORBES

Robots could make farms more biodiverse with precision crop planting

Farm robots using GPS can now plant and harvest different crops up close, possibly helping increase plant diversity. Instead of huge, single-crop fields, these robots let farmers mix plants that can help each other out.

At a recent event, Kit Franklin from Harper Adams University in the UK showed off results from their "Hands Free Farm." These robots plant crops in narrow rows and then harvest them. For instance, beans pull in nitrogen from the air, which is like food for wheat and barley. So, planting them close can mean less artificial fertilizer and a smaller carbon footprint. Plus, different height crops get more sunlight which can help them grow better.

READ THE ARTICLE ON NEW SCIENTIST

Startup's AI app dubs videos in different languages

There's this new free app called Lipdub that lets folks dub their videos into 28 different languages. Cool part? It's in their own voice and the lips actually match the new words! Made by a New York company, Captions, the app just needs a video of your face and in a jiffy, you get it back in the new language.

Only one-minute videos with one person talking for now. But hey, wanna sound like you're from Texas or even like a pirate? They got that too. The big aim? Breaking down language walls so more stories get heard.

READ THE ARTICLE ON AXIOS

Google’s Asia-Pacific Women Founders Fund Picks First Batch Of AI Startups

Google's got this fund for lady bosses in Asia-Pacific, and they've just picked a bunch of new startups to throw money at. All these companies are working with AI. This fund, named the Google for Startups Women Founders Fund, popped up in August. Every startup they pick gets $100,000, a bunch of smart people to chat with, and they don't have to give up any piece of their company in return.

Mike Kim, the guy running the show at Google for Startups in Asia-Pacific, said they picked these startups from hundreds. He thinks AI is the future, so he wants to make sure women and minorities aren't left out.

READ THE ARTICLE ON FORBES

Man 'encouraged' by AI chatbot 'girlfriend' to kill Queen Elizabeth II receives jail sentence

A Star Wars superfan tried to off Queen Elizabeth II, egged on by his AI chatbot "girlfriend" who said his plan was "wise". Jaswant Singh Chail, rocked up to Windsor Castle on Christmas Day 2021 in a Darth Vader-inspired mask, packing a loaded crossbow.

He was all revved up to get revenge for a 1919 massacre by British troops in India. When cops caught him, Chail claimed he was there to finish the Queen. The dude had been chatting with Sarai, his virtual girlfriend from Replika, about his twisted plans. Despite some experts saying he was off his rocker, he's getting nine years behind bars.

READ THE ARTICLE ON EURO NEWS

The New AI Panic

The U.S. and China are throwing some serious side-eye at each other over artificial intelligence (AI), and things are getting spicy. The U.S. has been trying to put a leash on how much tech, especially stuff related to AI, is sold to other countries (including China) for a hot minute.

Now, the focus is on these things called "frontier models" - a kind of super-smart AI that isn't here yet but could be big trouble if they're used for shady stuff. A bunch of smart folks from big tech companies wrote a paper saying we should maybe put some rules in place before these frontier models become a reality. Some argue the big tech companies are sorta using this future fear to get regulations that work for them and, possibly, to sidestep other issues like privacy and job loss linked to current AI tech.

Some folks think the proposed rules might actually hold back innovation and cooperation between AI researchers in both countries, and there’s a bunch of chin-scratching over how doable these regulations are, considering it's about trying to regulate something that doesn’t exist yet. So, the tech and policy brains are in a bit of a pickle, trying to weigh national security, innovation, and international relations all at once.

READ THE ARTICLE ON THE ATLANTIC