- The Intelligence Age
- 💬 Character AI Brings Multiple AIs To Your Group Chat
💬 Character AI Brings Multiple AIs To Your Group Chat
PLUS: Google’s AI electricity bill, a powerful new prompt technique, and more!
Welcome back, today we're covering how Airbnb’s CEO is thinking about the future of AI, Character.AI’s new feature, and Google’s growing AI electricity bill.
PLUS: A deep dive into step-by-step prompting, an AutoGen AI tutorial, and more.
Let's get started!
News & Insights
💬 Group AI Chats with Character.AI
The brainchild of former Google AI researchers, Character.AI now offers a group chat experience where users can interact with numerous AI characters simultaneously.
The Character Group Chat feature permits users to form a group chat with their preferred AI characters, their friends, or a blend of both. This means you can now create a social circle with friends, share ideas, and collaborate in real-time, just like in any typical group chat, but now with the AI companions thrown into the mix. techcrunch
Why does this matter?
Imagine the possibilities. A group chat with AI characters could potentially revolutionize brainstorming sessions in business environments, where diverse AI 'personalities' could contribute unique perspectives and ideas, thereby enriching the discussion.
🚀 Sidebar Is A Leadership Program That Accelerates Your Career
Once accepted, you’ll join your own personal Sidebar group of highly vetted peers, a professional coach, and custom programming to drive impact in your career.
Members join a virtual community with messaging and connect in fast-paced meetings to solve their biggest challenges at work, faster.
Employer reimbursements possible: many Sidebar members get their membership fees covered by their employer.
Why spend a decade finding your people - join the growing waitlist of top senior leaders, and apply to become a founding member.
🕹️ ‘If Only A Few People Control The Future, It’s Not Going To End Well’
Brian Chesky, Airbnb CEO, advocates for wide participation in AI's future. Chesky emphasizes the need for creative individuals to join the AI discourse, as the integrity of AI hinges on the diversity of its inputs. He equates artificial intelligence to driving at 1,000 miles an hour, with no option to reverse.
Products need to have a soul, Chesky suggests, and they should embody principles that benefit humanity. Referencing Apple's intersection of technology and humanities, he underlines the need for AI to foster positive human interactions. He believes that just like social media, AI could either isolate us in echo chambers or connect us in novel ways, depending on how it's used. fastcompany
⚡ Google’s AI Could Soon Consume As Much Electricity As Ireland, Study Finds
As AI comes of age, the energy toll begins to tell. A recent study reveals that if Google shifted its entire search business to AI, it would consume around 29.3 terawatt-hours annually, echoing Ireland's electricity consumption. That's a leap from Google's 2021 total electricity consumption, which stood at 18.3 TWh, with AI claiming a mere 10-15% of it. As the tech titan scales up its AI operations, with initiatives like the Bard chatbot and AI integration into its search engine, the energy implications are stark.
The study's scenario, however, relies on the assumption of full-scale AI adoption using current hardware and software. Alex de Vries, the author and academic at the VU Amsterdam School of Business and Economics, suggests that such a rapid transition is unlikely due to the limited supply of potent graphics processing units (GPUs) needed to handle the data deluge. So, while this scenario remains hypothetical for now, it illuminates a frequently overlooked aspect of the AI revolution: the spiraling energy costs of data centers, which already account for 1-1.3% of global electricity use. thenextweb
Why does this matter?
In the race to incorporate AI, let's not overlook the potential environmental trade-off. Could the next frontier for AI development lie in creating models that deliver optimal performance with minimal energy consumption?
🔨 Top Tools
🤖 Practical AI
What Else Is Happening
🗞️ More News
Microsoft disables the image generating feature of its Bing's AI system.
Startup Yepic breaks its promise, creating deepfakes without obtaining prior consent.
📈 Trending Today
Let's discuss Step-Back Prompting
Step-Back Prompting is like taking a step back to see the bigger picture before diving into the details.
It's based on the observation that we humans often simplify complex problems by first identifying the key, high-level concepts. We extract… twitter.com/i/web/status/1…
— Carlos E. Perez (@IntuitMachine)
Oct 11, 2023
🔭 Research Review
Neural Basis Models For Interpretability
Interpretability in AI is no longer a nice-to-have, it's a necessity, especially with the increasing deployment of AI in high-stakes domains. Addressing this, Meta AI has introduced a novel approach, a sub-family of generalized additive models (GAMs) dubbed Neural Basis Models (NBM). These models offer an intriguing blend of top-tier performance and glass-box interpretability.
NBM, a variant of GAMs, differentiates itself by learning a shape function for each feature. The beauty of this approach is that it allows for a clear understanding of a feature's impact on the prediction - simply visualize the shape functions. Interaction between variables is modeled by passing them into the same function, constructing a shape function based on that, a configuration known as GA2M. techcrunch
Written by Isaac R. Ward, Casey Clifton, and Alex Brogan.
Follow us for more AI content, including video news summaries, prompt ideas, business use cases, and more.