Welcome, AI enthusiasts

Nvidia is entering a new era of AI. At their annual GTC 2024 event, Nvidia has shared a variety of plans, and we've compiled them all for your convenience. In other news, Google’s Deepmind co-founder joins Microsoft in a shock move while AI mind-reading is getting better and better. Let's dive in!

In today’s insights:

• Nvidia’s Announcements During Their Annual GTC 2024 Event

• Google Deepmind co-founder joins Microsoft

• AI Mind-Reading Is Getting Better

Read time: 4 minutes

🗞️ LATEST DEVELOPMENTS

NVIDIA
🚀 Nvidia’s Announcements During Their Annual GTC 2024 Event

Evolving AI: Nvidia's GTC 2024 is all about generative AI - and the hardware that makes the current boom possible. They’ve announced so many new things that we’ll list them below in bullet points.

Details:

• Blackwell - Nvidia unveiled its new Blackwell platform, designed to enable generative AI with multi-billion parameter language models.

• Blackwell - The new generation of Blackwell GPUs is said to deliver four times the AI training performance, up to 25 times the energy efficiency, and up to 30 times the inference performance of the H100 GPU.

• Blackwell - This high inference performance applies to mixture-of-expert models such as GPT-4, while older models such as GPT-3 still run 8 times faster.

• Ceiba AI Supercomputer - Its AWS-hosted Ceiba AI supercomputer will now consist of 20,000 GB200 GPUs instead of the initially planned 16,000 H100 GPUs, showcasing a significant upgrade in performance and capabilities.

• Nvidia inference microservices (NIMs) - A new software platform designed to streamline the deployment of custom and pre-trained AI models into production environments. NIM will be available across AWS, Google Cloud, Microsoft Azure and Oracle Cloud marketplaces.

• Project GR00T - A foundation model for humanoid robot learning in simulation and the real world.

Why This Matters:

Nvidia's recent announcements have clearly set the stage for the future of AI. Considering that generative AI models need thousands of GPUs to operate, and with Nvidia holding over 80% of the GPU market, their impact is significant. During his keynote, Jensen Huang pointed out that they are now outpacing Moore's Law. This law, set by Intel co-founder Gordon Moore, predicted that the number of transistors on a chip would double roughly every two years—a prediction that has been accurate since 1965. However, Nvidia has achieved a X1000 improvement in the past 8 years. Realizing that most of today's AI models were trained on GPUs from the 2020 generation, we are heading towards an incredibly dynamic future in AI, where rapid advancements and the continuous pushing of boundaries are expected.

Lastly, with Project GR00T, Nvidia is set to power the humanoids of the future. Watch the 3-minute supercut below to see what’s coming.

AI: Balancing Risk and Return

Innovative technologies are revolutionizing business as we know it, and they’re more accessible than ever. But to truly harness the transformative potential of AI, you need to know how and when to use it. And which pitfalls to avoid.

The six-week Artificial Intelligence: Implications for Business Strategy online short course from MIT Sloan School of Management and MIT Computer Science and Artificial Intelligence Laboratory explores AI’s business applications and challenges. 

Choose this program to:

• Optimize your business: Leverage AI, ML, and robotics to drive efficiencies, improve productivity, and support your growth.

• Develop a strategic roadmap: Apply your knowledge to effectively integrate AI into your business.

• Gain a dual perspective: Benefit from a course designed by two prestigious schools — the MIT Sloan School of Management and the MIT CSAIL.

• Conveniently build career-critical skills: Follow a program that fits your schedule and benefit from 24/7 support and various payment options.

MICROSOFT
😮 Google Deepmind Co-Founder joins Microsoft

Evolving AI: Microsoft hires DeepMind Co-Founder Mustafa Suleyman as CEO of new Microsoft AI devision.

Key Points:

• Suleyman to lead new consumer AI division, including Copilot, Bing, and Edge.

• Addition of Inflection AI's talents, including co-founder Karén Simonyan.

Details:

Microsoft has appointed Mustafa Suleyman, co-founder of Google DeepMind, as CEO of its new AI division. This division will steer the development of consumer-facing AI products such as Copilot, Bing, and Edge. Suleyman's role includes joining the senior leadership team, underlining his influential position in shaping Microsoft's AI strategy. His appointment comes alongside Microsoft's continuous investment in AI, including partnerships with OpenAI and the recruitment of talent from Inflection AI, signaling a potent augmentation of its AI capabilities and ambitions.

Our Thoughts:

This move not only leverages Suleyman's expertise but also signals Microsoft's aggressive pursuit of AI innovation. However, it raises the question of what the future holds for Inflection AI, as notable talents have now moved to Microsoft. As one of the most prominent AI startups, Inflection AI was founded in 2022 by LinkedIn co-founder Reid Hoffman and Suleyman. The company aims to develop AI systems capable of engaging in open-ended dialogues, answering questions, and assisting with a variety of tasks. Inflection AI continues with a new CEO and is bringing its models to the Azure cloud, including the recently released Inflection 2.5.

MINDEYE2
🧠 AI Mind-Reading Is Getting Better

Evolving AI: A breakthrough in visualizing thoughts.

Key Points:

• A leap in reconstructing visuals from brain activity.

• Groundbreaking method uses just 1 hour of fMRI data.

• Novel approach enhances image retrieval and reconstruction.

Details:

The field of visual reconstruction from brain activity has historically faced limitations due to the need for extensive, subject-specific training data. This latest research, however, introduces a game-changer: high-quality visual reconstructions with merely one hour of fMRI training data. This innovation comes from a method that combines pretraining on multiple subjects with a unique fine-tuning process for new subjects. The technique employs a functional alignment procedure that maps brain activity to a common latent space and then transitions into image space, dramatically reducing the need for extensive training while maintaining precision in visual reconstructions.

Why It Matters:

The advancements in reconstructing visual perceptions from brain activity mark a significant milestone in neuroimaging and AI integration. This breakthrough not only makes the technology more accessible by reducing the time and cost involved but also enhances the quality of reconstructions. It paves the way for practical applications such as aiding in communication for those unable to speak or enhancing virtual reality experiences. This research could fundamentally change our approach to understanding and interacting with human cognition, blurring the lines between thought and visual representation.

💡 Tip of the Day

Stability AI introduced their latest model for rendering 3D videos. This advanced functionality is now accessible for commercial purposes via a Stability AI Professional Membership, priced at $20 per month for creators and developers generating less than $1 million in annual revenue.

🚀 Nvidia: How they help scaling quantum computing. 🚀 

👋 Gemini: Everything you need to know about Google’s AI platform.

🤖 Sam Altman: OpenAI CEO Sam Altman promises 'amazing' AI model for 2024

📈 Trending AI Tools

• 🔊 Recast - Convert articles into audio summaries (link)

• 📜 Holler: AI-powered survey tool to collect feedback and insights (link)

• 🎭️ FaceVary: A tool to swap faces using customizable templates for diverse images (link)

• 🎥 Pika Labs: A tool to generate videos from prompts (link)

• 🎧 Listnr - AI audio tool encompassing text-to-speech, speech-to-text, and voice cloning features (link)

• 🎨 StarryAI - A prominent AI image generation tool on the internet landscape (link)