#101. 🤖 Google Duet AI available 🏞️ ID'ing AI-generated images ⚠️ $400 to build an AI disinformation machine
PLUS: Role-playing games in ChatGPT ✨ AI-powered SEO optimization ✨ AI how-to guides
Marc Hoag
August 30, 2023
❓Is this new format of Future Perfect too short now?Instead of 9 article summaries per issue it's just 3 now. Is that ok for you? |
📰 AI NEWS
Google rolls out Duet AI across its Workspace apps, aiming to serve as a productivity and collaborative tool for tasks like generating emails, slides, and charts.
Duet offers app-specific features, such as AI-based lighting in Google Meet and automated summaries in Google Chat.
The service costs $30 per user for large organizations, comparable to Microsoft's Copilot, which offers similar functionalities across Office apps.
While Duet strives to be accurate, users are cautioned to double-check its outputs, especially for business-critical data, due to the imperfections of the underlying AI models.
SynthID is a new tool developed in partnership with Google Cloud to watermark and identify AI-generated images, particularly those created by Imagen.
The technology embeds an imperceptible watermark into the pixels of an image for easy identification, aiming to curb the spread of misinformation and build trust.
While not foolproof, SynthID is effective against many common image manipulations and works in tandem with traditional identification methods like metadata.
The tool is in its beta phase and available to a limited number of Vertex AI customers; it may evolve to work with other types of AI-generated content such as audio and video.
A developer known as Nea Paw used OpenAI's text generation technology and other AI tools to create a disinformation campaign against Russian state messaging for around $400. The campaign, called CounterCloud, wasn't publicly posted but shared with WIRED.
The project aims to expose the ease and low cost of using AI for mass-produced disinformation. Paw, who claims to be a cybersecurity professional, says there's no easy solution to this growing issue.
Experts warn that AI could be increasingly used for crafting highly personalized propaganda and interacting with social media users. Some instances of AI-powered disinformation and its use in legitimate political campaigns have already been observed.
OpenAI and other organizations are concerned about the misuse of their technology. However, the challenge remains in educating the public and finding effective methods to block or detect AI-generated disinformation.
⚖️ AI LAW FROM THE MARC HOAG LAW BLOG
Sorry, nothing new for today; please check back again tomorrow!
⚡️ CHATGPT POWER PROMPTING!
So this time I had some fun with a remarkably simple yet incredibly powerful prompt designed to evoke the old text-based role playing games of back in the day. Some of you may be too young to even know what these were, in which case, you’re in for a treat! Think of them as sort of "interactive books.”
Let's play a text-based role-playing game where you describe scenes to me, give me options for how to proceed, and so on. Let's set it in the context of [insert your favorite fictional universe].💡 AI RESOURCES
Mango SEO: AI-powered SEO optimization.
Prompt Pal: ChatGPT prompts.
Guidde: AI-generated how-to guides, etc.
🎉 THAT’S ALL FOR TODAY!
What'd you think of today's email?This helps me improve it for you! |
🙏 PLEASE SHARE this post if you liked it!
Thanks for reading!
-Marc 👋
Looking for past newsletters? You can find them all here.