- Future Tools
- Posts
- Who's winning the multimodal AI race?
Who's winning the multimodal AI race?
And the fastest AI system to-date
What a week it’s been…starting with a big shake-up in Big AI leadership. Anthropic snagged Instagram co-founder Mike Krieger, while OpenAI officially said goodbye to Chief Scientist Ilya Sutskever (after a year of tensions).
What does this mean for these companies? I say Anthropic is gearing up to strengthen its mobile app team with seasoned consumer tech veterans. Meanwhile, OpenAI’s loss of Ilya signals some final leadership restructuring after Sam Altman’s rapid firing and reinstatement last year.
Both companies are positioning themselves for the next leap in AI—let’s see who makes the bigger splash!
OpenAI's Biggest News: GPT-4o and Custom GPT Store
OpenAI
OpenAI's Spring announcement had everyone on the edge of their seats. Speculations swirled around new search features or even GPT-5. Instead, groundbreaking model updates with GPT-4o and the newly accessible Custom GPT Store stole the show.
First up: GPT-4o, a natively multimodal AI marvel, can process text, audio, and images all at once—and it can even talk. This new model promises a more natural human-computer interaction, like talking to a knowledgeable friend or personal assistant.
A major upgrade. GPT-4o is 2x faster, half the price (for developers using API), and smarter than its predecessor, with significant improvements in understanding vision and audio. It’s a leap forward in AI capability, offering real-time responses across multiple modes of input.
Imagine conversing with an AI that responds in milliseconds, almost like a human. This speed, combined with enhanced multilingual capabilities, makes GPT-4o a versatile tool for global applications.
The sky is the limit. GPT-4o is available to everyone (no premium subscription needed), from individual users to large enterprises, through OpenAI's platform. You can use it for all your generative AI needs, including customer service bots, content creation, personalized education tools, software development, and more.
Next: OpenAI’s Custom GPT Store. Originally launched as a subscription-only feature, the GPT Store is now free for all. This store is a treasure trove of GPTs—custom chatbots made by ChatGPT Plus subscribers designed to cater to specific needs.
Some popular GPTs include Consensus, which helps with scientific research, and a logo-making bot. I cover more GPTs in this video!
Accessible to all. In January, the store was only accessible to paid subscribers. Now, anyone can browse and use these bots, democratizing access to powerful AI tools.
Encouraging innovation. With an engagement-based revenue-sharing program for GPT builders, the platform encourages innovation and creativity. Users can now explore a myriad of bots without the paywall, expanding the reach and impact of these AI solutions.
Why it matters: OpenAI’s latest moves signal a strategic shift towards inclusivity and accessibility. These updates underscore OpenAI's vision of integrating AI deeply into our daily lives, making advanced technology available to a broader audience.
Google I/O Unveils Project Astra and "Web" Search Filter
Google I/O Conference
Google I/O conference was loaded with AI announcements. Highlights included Project Astra, a groundbreaking multimodal AI assistant, and a new "Web" filter for classic text-based search results, balancing innovation with user needs.
First up: Project Astra, a multimodal AI assistant that could transform into a real-life Jarvis. This model, poised as a Siri competitor, can see and react to the world around it. You can talk, type, draw, snap pics, or shoot videos to chat with it. But Astra is not just about chatting—it's about getting stuff done:
Astra's underlying tech is powered by the Gemini 1.5 Pro, which boasts a massive context window. Think of it as giving the AI a superhuman memory.
Google has been working hard on speed and latency to make this AI snappier than ever.
While Astra is still in its early prototype phase, it’s already generating buzz as the next big thing in AI assistants.
Next: Google's new "Web" filter. This feature lets users filter for those classic text-based links (you know…the blue ones) before AI summaries soon become the norm. This feature shows just how many users are sticking to legacy interfaces—AKA the good, old-fashioned web pages.
This feature is rolling out globally, sparking debates in the SEO world about the future of search.
Why it matters: Google's moves with Project Astra and the "Web" filter signify a balanced approach to AI and user needs.
Astra showcases how AI can integrate seamlessly into daily life, acting as a true assistant rather than just a chatbot. Meanwhile, the "Web" filter respects some users’ desire for classic search results, proving that Google values diverse ways of accessing information.
Turn AI Into Your Personal Assistant
Discover how to turn AI into your personal productivity powerhouse with HubSpot’s highly anticipated AI Task Delegation Playbook. Master the art of AI delegation and optimize your workflow like never before. Get ready to save time and boost efficiency with their easy-to-use templates and calculators. Don’t miss out—download your copy today and start transforming your workday.
Leverage AI for streamlined task management, significantly enhancing time efficiency
Utilize AI tools to elevate decision-making and maximize workflow efficiency across teams
Explore comprehensive templates and detailed examples to master straightforward AI delegation
Evaluate and optimize productivity by assessing measurable impacts of AI on your daily output
Intel’s Aurora Becomes Fastest AI System
This Monday, Intel blew past the exascale computing threshold with their new supercomputer, Aurora. This is now the fastest AI-focused system on the planet.
What’s exascale computing? It’s a measure of performance that’s a quintillion calculations per second (yeah, that’s a lot).
What makes it so fast? In short: the largest GPU Cluster ever. Aurora is built on the Intel Data Center GPU Max Series and the innovative Intel Xe GPU architecture. This setup allows for insane parallel processing capabilities, essential for crunching complex neural network computations.
Picture this: 166 racks, 10,624 compute blades, 21,248 Intel Xeon CPUs, and a whopping 63,744 Intel Data Center GPUs. That’s Aurora.
Advancing scientific discovery. It’s not just about speed. This AI-centric supercomputer is designed to handle generative AI models, which means it’s set to accelerate scientific discovery like never before. From mapping the human brain’s 80 billion neurons to enhancing particle physics and speeding up drug design, Aurora is the new frontier in AI research.
Why it matters: This isn’t just a win for Intel; it’s a leap forward for science and humanity. Aurora's achievement sets a new benchmark in AI and high-performance computing, pushing the boundaries of what's possible—bound to drive innovation across various industries.
Everyone in Hollywood is using AI, but "they are scared to admit it."
Cisco Hypershield revamps cybersecurity with AI-powered defense.
Google I/O 2024: 100 things Google announced.
Senate study proposes ‘at least’ $32B yearly for AI programs.
Hugging Face is sharing compute to help beat the big AI companies.
Sony Music warns tech companies over ‘unauthorized’ use of its content to train AI.
More important AI news: Dive deeper into this week’s hottest AI news stories (because yes, there are even more) in my latest YouTube video:
How is AI reshaping our world? On the latest episode of The Next Wave, we dive into AI's transformative potential. We explore revolutionary impacts on societal structures, healthcare advancements, streamlined government efficiency, and tackling global challenges like climate change and hunger.
I don’t know about you, but I think that’s enough groundbreaking updates for this week—time to take a breather and watch how these giants shape the future!
Drop me a line with your favorite model update from this past week.
—Matt (FutureTools.io)
P.S. This newsletter is 100% written by a human. Okay, maybe 96%.