OpenAI fires co-founder and CEO Sam Altman

☕️ Happy Friday, and welcome to my daily dive into the Generative AI landscape.

Today’s issue includes:

• OpenAI fires co-founder and CEO Sam Altman

• Foundational models

• Introducing DeciLM

• Feedback from Copilot’s Earliest Users

TOP STORIES 

♨️ Sam Altman, CEO of OpenAI, has been dismissed from his position, as announced by the company.

The decision followed a review by the board, which found Altman's communication lacking in candor, undermining their governance. Mira Murati, the Chief Technology Officer, steps in as interim CEO, with a permanent successor search underway. OpenAI's staff, like the wider public, were informed of this development simultaneously.

Altman, on his X (formerly Twitter) account, reflected positively on his tenure at OpenAI. This departure is particularly startling given Altman's prominent role in OpenAI, especially after the successful launch of ChatGPT and his recent participation in key events, like the DevDay conference and the Asia-Pacific Economic Cooperation conference.

Microsoft, a significant investor in OpenAI, reaffirmed its commitment to the partnership under Murati's leadership. Altman has also been involved in discussions about creating an "iPhone of artificial intelligence" and is a major stakeholder in Humane, which is launching its Humane AI Pin.

Furthermore, OpenAI announced that co-founder Greg Brockman is stepping down as chairman of the board but will remain with the company. Altman, a co-founder, initially co-chaired with Elon Musk, who left in 2018 to avoid conflicts with Tesla and later started xAI, another AI venture.

♨️ Foundational models have become linchpins in the evolution of generative AI, a phenomenon reshaping our technological landscape. These models are expansive neural networks, trained on vast data troves, capable of learning a range of skills without specific task-oriented training. This versatility stands as their hallmark; a single foundational model can perform text generation, image creation, and even complex problem-solving, adapting to diverse applications with remarkable agility.

Generative AI, on the other hand, leverages this versatility to create new content. From art to literature, it synthesizes original works that are indistinguishable from human-created content. This capability has profound implications. For professionals and creatives, it's a tool that amplifies productivity and innovation. Yet, it also stirs ethical and practical debates, from authorship rights to the potential for misuse.

Together, foundational models and generative AI are not just tools but catalysts of a paradigm shift. They challenge our traditional notions of creativity and intellectual output, urging us to rethink the boundaries between human and machine capabilities. As they evolve, they promise to unlock new horizons in technology and creativity, while posing questions that will shape the future of AI ethics and governance.

♨️ Introducing DeciLM: DeciLM 6B, a new Large Language Model (LLM) with 5.7 billion parameters, introduces remarkable efficiency in AI modeling. It achieves 15 times higher throughput than Llama 2 7B while maintaining quality, ranking high among 7 billion parameter LLMs. DeciLM’s unique architecture, developed using Deci’s AutoNAC Neural Architecture Search engine, features innovative Grouped-Query Attention, enhancing computational and memory efficiency. The model, available under a permissive license, is poised to advance generative AI applications with reduced costs and environmental impact. DeciLM 6B, coupled with Deci’s Infery-LLM inference SDK, marks a significant step in efficient, scalable AI technology.

♨️ Feedback from Copilot’s Earliest Users:

• Introduction of Copilot for Microsoft 365: Launched eight months ago to reduce digital clutter and enhance productivity.

• Generative AI's Potential: Focus on its ability to transform work.

• Productivity Gains Confirmed: New data shows significant improvements in efficiency and creativity for Copilot users.

• User Feedback: High satisfaction, with 77% of early users reluctant to work without Copilot.

• Organizational Impact: Beyond personal productivity, Copilot is seen as elevating the entire organization's capabilities in writing, design, coding, data analysis, etc.

• Amplifying Expertise: Helps take work from good to exceptional.

• Research Findings:

  • 70% of users feel more productive.

  • Users are 29% faster in tasks like searching, writing, and summarizing.

  • Copilot helps catch up on missed meetings 4x faster.

  • 64% spend less time on email.

  • 85% reach a good first draft faster.

  • 75% save time finding information in files.

• Preference Over Free Lunch: Users prefer Copilot access over free lunch at work.

• Impact on Specific Roles:

  • Sales: Users save 90 minutes a week; 83% feel more productive.

  • Customer Service: 12% reduction in case resolution time.

  • Security: 44% more accuracy and 26% faster task completion.

• Business Process Transformation with AI: Survey indicates AI can fundamentally transform business processes, with generative AI being particularly effective in information retrieval.

• 

IN THE NEWS

• Gen AI: A guide for CFOs: CFOs must proactively adopt generative AI, focusing on valuable projects and managing risks. This technology offers significant opportunities for value creation, necessitating strategic resource allocation by CFOs at both enterprise and finance function levels. They should prioritize impactful use cases, comprehend the technology, and foster innovation. Engaging with generative AI is crucial for CFOs to spur growth, efficiency, and competitive edge, preparing their teams for future challenges.

• Hype not hyperintelligence is lifting values: This article highlights the growing investment in Generative Artificial Intelligence (GAI) amid skepticism about its limitations and ability to achieve human-like intelligence. It discusses the challenges GAI faces in understanding new scenarios and error identification, while also noting its potential applications and impact on corporate profits.

• Introducing Emu Video: Meta's new generative AI advancements include Emu Video, which efficiently generates high-quality videos from text using diffusion models, and Emu Edit, offering precise image editing with unparalleled control and accuracy. These technologies exemplify significant progress in AI-powered creative tools, enhancing digital expression and personalization.

• AI Model for Music: Google DeepMind, in collaboration with YouTube, introduces Lyria, an advanced AI music generation model, and two new AI experiments: Dream Track and Music AI tools. These initiatives aim to revolutionize music creation, offering innovative ways for artists to interact with fans and enhance their creative processes, marking a significant leap in AI's role in the music industry.

• 10 HR Trends in 2024: In 2024, HR trends focus on diversity, equity, inclusion, generative AI, compliance, and talent. Key aspects include prioritizing inclusion, proactive DE&I measures, pay transparency, ethical use of AI, intelligent HR tech upgrades, employee wellness, remote worker compliance, state-specific benefits, skill development, and evolving career paths. These trends are reshaping HR and leadership strategies in a dynamically changing workplace.

TIPS & TOOLS

Leverage AI to grow on LinkedIn: Get more reach, engagement and followers in 10min per day.

Join my community by subscribing to my newsletter below:

My Community

Join my professional communities on LinkedIn