- NATURAL 20
- Posts
- Google's Game-Changing Approach to Content Creation
Google's Game-Changing Approach to Content Creation
Discover How Quality Takes Center Stage and the Intricacies of Labeling AI Generated Content, Revealed at Search Central Live Tokyo 2023
![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/151875c2-789e-4e16-b5e7-1737bdf68f75/natural_20_banner.png)
The AI Verdict
A newsletter that tracks regulations on AI and focuses on AI's impact on legal professionals. Also highlights emerging AI tech.
Today:
Google Is Forming A Policy On AI Content And E-E-A-T
Looks like Google's gettin' deep in the weeds about AI-generated content. Our pal Gary Illyes and co. from Google decided to spill the beans at Google Search Central Live Tokyo 2023, and Kenichi Suzuki was good enough to jot down the highlights.
Google doesn't give a hoot whether your content is spat out by a computer or a person. What it really cares about is the quality of that content. Turns out, Google doesn't really distinguish between AI and human-generated content. They don't label it, and they ain't asking publishers to label theirs either. Even with the EU asking folks to label AI content to fight fake news, Google's taking a backseat. They figure it's up to publishers to decide if they wanna play that game.
But before you get all giddy and start firing your writers, Google's got a word of caution. They advise against publishing AI content as-is, without a human editor giving it the once-over. Same goes for translated content. Also, remember that Google's algorithms are designed to favor natural content. The more human-like the content, the higher it's likely to rank.
How does Google deal with AI content when it comes to E-E-A-T? That stands for Experience, Expertise, Authoritativeness, and Trustworthiness, in case you're wondering.
Mosyle brings generative AI to Apple mobile device management
Mosyle, a mobile device management (MDM) vendor, announced it's adding a sprinkle of generative AI to help handle, secure, and ensure compliance for Apple's macOS devices. All part of the ongoing feast of the Mosyle Apple Unified Platform, funded by a jaw-dropping $196 million.
Mosyle CEO, Alcyr Araujo, confessed that their goal is to give customers an easy access to the world of scripting, which can automate a whole lot of things. With a hint of irony, it seems Araujo, doubling as Mosyle's IT admin, got this idea while creating a script for macOS. He figured that using gen AI with their script catalog project could let users whip up a script or find one by simply asking.
Mosyle's new tool, AIscript, leans on OpenAI's GPT models, but Araujo is keeping his options open. The company isn't just hooking OpenAI's API to its MDM tech, mind you. They've gone the extra mile to protect user data and make sure the generated scripts aren't a bunch of mumbo-jumbo.
China's ByteDance Has Gobbled Up $1 Billion of Nvidia GPUs for AI This Year
Chinese tech honchos ByteDance, Tencent, Baidu, and Alibaba are buying up Nvidia's tech like they're shopping at a Black Friday sale. They're especially hooked on Nvidia's high-performance computing stuff.
According to a report, ByteDance, the guys behind TikTok, have matched what the whole of China spent on Nvidia gear in 2022. That's about $1 billion worth of tech in 2023 alone, and the year's only halfway through.
ByteDance splurged on Nvidia's A100 and H800 cards, even after Uncle Sam told Nvidia to stop selling its top-of-the-line goods to China back in August 2022. After some grumbling about losing $400 million in revenue, Nvidia played nice and made a special H800 card that could be exported without ruffling the government's feathers.
Hidden Door launches playtests for Wizard of Oz games created by generative AI
Hidden Door, a game tech studio in the Big Apple, is in the works to launch a "Wizard of Oz" role-playing game, using generative AI to customize each adventure. They're leveraging the public domain work of L. Frank Baum, providing players with unique narratives within the "Wizard of Oz" universe – join the Wicked Witch, wander across Oz, or just stay put in Kansas and chill with Toto.
Playtesting will be conducted on Discord, with a full launch anticipated by winter 2023. The game adapts to player decisions, providing an exclusive narrative for each adventure. It's like having a choose-your-own-adventure book, but with AI spinning the yarn.
The company plans to use this approach with other intellectual properties in the future. So, the Potterheads, Trekkies, and Middle Earth junkies might have some exciting times ahead! They're essentially turning cherished works of fiction into tabletop-esque RPGs powered by AI.
AI-guided Remediation for IaC Security / KICS
Infrastructure as Code (IaC), a method businesses use to manage their digital groundwork, kinda like making a blueprint for your online estate. But like any good tool, it's only as sharp as the person wielding it. Slip-ups, or 'misconfigurations,' in setting up this infrastructure can lead to some nasty bumps in the road - think security loopholes, operational kerfuffles, even breaches.
These blunders can make the life of a dev team feel like they've bitten off more than they can chew, and not just in the security sense. Troubleshooting these issues can turn into a real wild goose chase. The upshot? Longer deployment cycles, more room for errors, and a downright headache for the folks in charge of keeping things running smoothly.
AI Guided Remediation for IaC Security and KICS by Checkmarx. This AI helper not only shows you the problems, but gives you the skinny on how to fix 'em too. Plus, it's got a keen eye for any secret info that might accidentally get spilled. It's the proverbial watchdog for your code's secrets, ensuring they don't get loose and cause a hullabaloo.
Parallel Domain’s API lets customers use generative AI to build synthetic datasets
San Fran's hotshot startup Parallel Domain is handing out the keys to its virtual world-building workshop. Their brand spanking new API, dubbed Data Lab, lets users whip up synthetic datasets - a.k.a. fake but useful data - using generative AI. It's as easy as pie - download it from GitHub, jot down some Python code, and you're off to the races.
![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/46aa730f-d749-4b18-a601-152db83b285f/unleash-innovation-desktop.jpg)
This powerful tool's main aim? To give autonomy, drone, and robotics firms more say and speed in cooking up big datasets. Basically, it cuts the waiting time for a fresh batch of data to as quickly as the user can type up a request. The result? Quicker, deeper learning for AI models.
Major carmakers and self-driving companies are already biting. Before, it could take weeks or even months to generate specific datasets. Now, with Data Lab, it's almost immediate.
But it doesn't stop at self-driving tech. Parallel Domain's sights are set on other sectors where computer vision can work its magic, like agriculture, retail, or manufacturing. The founder and CEO, Kevin McNamara, envisions a future where if an AI needs to learn to see the world, it starts with Parallel Domain.
Ethereum Scanner Etherscan Adds OpenAI-Based Tool to Analyze Smart Contract Source Code
Etherscan, a top-dog Ethereum blockchain watchdog site, whipped up a new toy that lets folks make heads or tails of smart contract source code, all with the help of artificial intelligence. They built it with OpenAI's tech, so users can ask this tool - called Code Reader - to explain the nitty-gritty of these smart contracts, whether it's the whole kit and caboodle or just bits and pieces. And hey, it'll even show you how to read and write these contracts, so you can feel all smart and savvy when you're working with decentralized apps.
Etherscan gave a heads-up that the tool is just for your info. To use Code Reader, you gotta hook up to OpenAI's API and make sure you've got enough usage limits. API's just tech talk for a bridge that lets two computer programs gab and swap notes. Right now, Code Reader's a bit of a chatterbox - it doesn't do back-and-forths, it's strictly one-off prompts.
The World's First AI DJ Has Officially Hit Radio Airwaves
A radio station in Oregon, KBFF Live 95.5 FM, has written a new page in the history books by launching the world's first AI (that's fancy talk for 'Artificial Intelligence') DJ. And no, it's not a robot spinning records, but a software named RadioGPT. It's kind of like a digital twin of their regular host, Ashley Elzinga.
We made history as the world's first radio station with an AI DJ! Our midday host Ashley has become AI Ashley! We can’t wait for you to meet Ashley, the world's first artificially intelligent DJ. As to the intelligence of our other DJ’s…we’ll save that for another post ;) twitter.com/i/web/status/1…
— Live 95.5 (@live955)
6:29 PM • Jun 13, 2023
The synthetic Ashley sounds so real, it had a gab fest with listeners and even handed out Taylor Swift tickets. And the real Ashley? Well, she was probably kicking back with a soda, joking about having the day off.
What'd you think of today's edition? |
What are MOST interested in learning about AI?What stories or resources will be most interesting for you to hear about? |
Reply