- /IMAGINE
- Posts
- Google's Gemini AI Goes Off The Rails! 🤦🏻♂️
Google's Gemini AI Goes Off The Rails! 🤦🏻♂️
And → More Google news, Stable Diffusion 3, prompts, tutorial, and more!
Welcome back, my AI-obsessed friends.
Shake off the sleep and have your coffee—it’s Friday. It was an interesting day in the AI world! This is the latest issue of /Imagine, your go-to source for the freshest news and developments in AI. Here’s EVERYTHING you need to know. Let’s dive in…
🛑 Google Gemini’s Pause: Their Image Generation Went Off The Rails
The Bytes:
✣ Google has some history homework to do. They halted the image generation of people in Gemini to address race accuracy issues, following widespread criticism and concern.
✣ The AI community on Twitter expressed significant uproar over the model’s radically inaccurate depictions. It’s so bad it’s become a meme.
Gemini is just so messed up.
— Wall Street Silver (@WallStreetSilv)
1:21 AM • Feb 22, 2024
✣ Google has committed to fixing these inaccuracies but it continues to illustrate the tech industry’s ongoing struggle with AI bias and representation.
Why you should care: We have created a powerful tool, and one of the first things we did with it was make it lie and alter reality. Google’s temporary suspension of Gemini’s image generation feature is a stark reminder of the challenges centralized AI faces. The community’s reaction underscores the importance of getting it right. As users and creators, we rely on these tools to be fair and accurate. I don’t want to bash them too hard. The fact that they are issuing such a quick response is a positive sign that the industry is willing to take the necessary steps to improve AI responsibly. Remember folks, it’s always good to own and run your AI locally.
🎨 Stability AI Drops Stable Diffusion 3
Image Credit: Stability AI
The Bytes:
✣ Stability AI just pulled back the curtain on Stable Diffusion 3, an early preview of their most advanced text-to-image model yet.
✣ The model boasts improvements in multi-subject prompts, image quality, and word abilities, setting a new standard for AI creativity. It even uses a new Sora-like diffusion transformer architecture.
✣ Whether you are an individual creator or an enterprise, you are covered with models ranging from 800M to 8B parameters.
Why you should care: Stable Diffusion 3 isn’t just an incremental upgrade; it’s a leap forward in the democratization of AI-generated art. Whether you’re a hobbyist or a professional, this tool promises to unlock new realms of creativity and expression. To get on the waitlist, head here.
💡 Mystery AI Company Revealed: Google’s $60M Reddit Data Deal
Image Credit: Reddit
The Bytes:
✣ This week, we wrote about a mystery company that was paying for Reddit’s vast content for $60 million annually, aiming to refine its AI models. We learned today that the company was Google.
✣ Reddit’s upcoming IPO is buoyed by this news, with plans to list under the ticker RDDT. Fun fact: Sam Altman of OpenAI holds an 8.7% stake in Reddit, and stands to gain significantly from the IPO.
✣ The deal grants Google a pass to Reddit’s Data API, enhancing the search engine’s understanding of user-generated content.
✣ Some Redditors are raising an eyebrow over the deal, citing concerns over data privacy. That being said, the agreement could lead to more human-level AI systems.
Why you should care: Google’s investment in Reddit’s data is a significant move for AI development. It means that Google’s AI could become more nuanced and context-aware, potentially transforming how we interact with search engines and AI assistants. However, it also raises questions about data privacy and the ethical use of user-generated content. As AI becomes more integrated into our digital lives, the balance between innovation and privacy remains a critical conversation.
Image Credit: User @morea_r on Midjourney
Prompt Share: Cinematic, Picture Mona Lisa at a Starbucks window, dressed in a modernized variation of Steve Jobs' iconic attire, complete with similar glasses. Seated at a desk with a laptop and a coffee, capture her engaged in creative work, with the original Mona Lisa background visible outside, blending classical art and contemporary tech vibes --ar 9:16
Check out my Midjourney starter kit for more tips here!
Join Us: Don't forget to follow me on Twitter/X at @jake_joseph and the brand on its fresh new page @slashimagineai. Let’s be friends.
Share The Insights: Know someone else who's obsessed with AI? Forward this newsletter and let's expand the community.
Feedback
How did we do with today's newsletter?Share your vote to help us deliver the #1 resource on AI for our community. |
Your perspectives help shape the content and hopefully help me make the newsletter suck less every day. Share your feedback on this issue or what you'd like to see next by directly replying to this email or by reacting to the poll above.