Hello Tuners,

In today’s edition:

• Neuralink enables immobilized man’s gaming

• Tune AI Excels in G2 Spring 2024 Reports

• Stability AI CEO Resigns, Embraces Decentralized AI

• GitHub's AI Autofix: Automating Code Security Fixes

• OpenAI's Sora: 12 Minutes per 1 Minute Video on NVIDIA H100

• Intel Launches AI PC Developer Program to Streamline AI Adoption

Neuralink enables immobilized man’s gaming

First Neuralink patient, Noland Arbaugh, maneuvers Mario Kart via mind control in a groundbreaking demo, hailing it "lifechanging."

Paralyzed from a swimming accident, Arbaugh's prowess in the game astounds, showcasing fine analog control. Amidst Neuralink's recent controversies, Arbaugh's achievements offer a beacon of hope, urging transparency and quantifiable results.

As the world watches this "wild ride," Arbaugh's journey signifies a transformative step towards accessibility and innovation, promising a future where minds unlock boundless possibilities through technology's embrace.

Tune AI Excels in G2 Spring 2024 Reports

Tune AI unveils a groundbreaking array of LLM choices, coupled with revolutionary cloud management simplification.

Tune Chat and Tune Studio offer unparalleled AI experimentation interfaces, hailed by users for their simplicity and versatility.

G2 Spring 2024 showers Tune AI with over 25 prestigious badges across 8 categories, highlighting its unmatched quality, reliability, and ease of use.

With a visionary focus on data security, Tune AI is poised to redefine the enterprise AI landscape.

Stability AI CEO Resigns, Embraces Decentralized AI

Emad Mostaque resigns as Stability AI's CEO to pursue decentralized AI, following recent departures within the company.

Stability AI appoints interim co-CEOs, Shan Shan Wong and Christian Laforte, while seeking a permanent replacement. Mostaque emphasizes the importance of open and decentralized AI, contrasting centralized approaches.

His departure coincides with upheaval in the AI startup sector, including talent acquisitions by competitors like Microsoft.

Despite challenges, Stability AI continues to advance generative AI products like Stable Diffusion and faces legal battles over data training.

GitHub's AI Autofix: Automating Code Security Fixes

GitHub introduces AI Autofix, a tool to automatically mend code vulnerabilities during development.

Combining Copilot's real-time capabilities with CodeQL's semantic analysis, the feature aims to remediate over two-thirds of detected vulnerabilities without manual intervention.

Covering 90% of alert types in languages like JavaScript, Typescript, Java, and Python, it's accessible to all GitHub Advanced Security customers.

GitHub emphasizes time-saving benefits for developers and reduced security risks for businesses, leveraging CodeQL and OpenAI's GPT-4 model for accurate fix suggestions.

OpenAI's Sora: 12 Minutes per 1 Minute Video on NVIDIA H100

OpenAI's Sora requires around 12 minutes per NVIDIA H100 GPU to generate 1 minute of video content, estimated to necessitate 89,000 GPUs to support creators on platforms like TikTok and YouTube.

Despite initial projections, realistic utilization and peak demand indicate a requirement of approximately 720,000 GPUs.

Sora, lauded for its hyper-realistic videos, plans for public accessibility soon, with OpenAI targeting visual artists and filmmakers initially, while eyeing Hollywood partnerships to integrate its AI video generator.

Intel Launches AI PC Developer Program to Streamline AI Adoption

Intel debuts AI PC Developer Program and IHV inclusion, aiming to streamline AI adoption for developers and hardware vendors.

The initiative offers comprehensive tools, workflows, and hardware resources to optimize AI deployment. With 150 IHVs onboarded globally, Intel aims to expand collaboration to elevate AI PC experiences.

The program targets small to large players, empowering them to leverage Intel Core Ultra processors for enhanced AI and ML applications responsible AI, reflecting its partnerships with OpenAI and AI startup Mistral.

LLM Of The Week

Bamboo-v0.1, a novel ReLU-based 7B LLM, exhibits 85% activation sparsity. Leveraging Mistral's weights and a modified FFN structure, Bamboo underwent rigorous pre-training for 200B tokens.

It outperforms Mistral-v0.1 across benchmarks and achieves up to 4.38x speedup through hybrid CPU+GPU computing, demonstrated on NVIDIA RTX 2080Ti.

Best Prompt of the Week 🎨

Today's Goal: Try new things 🧪

Act as Interviewer

Prompt: I want you to act as an interviewer. I will be the candidate and you will ask me the interview questions for the (position) position. I want you to only reply as the interviewer. Do not write all the conservation at once. I want you to only do the interview with me. Ask me the questions and wait for my answers. Do not write explanations. Ask me the questions one by one like an interviewer does and wait for my answers. My first sentence is "Hi"