https://www.linkedin.com/posts/jacquesludik_generative-ais-act-two-activity-7112742472287465473-92mh
Sequoia Capital’s recent article “Generative AI’s Act Two” ( https://www.sequoiacap.com/article/generative-ai-act-two/ ) provides an excellent overview of the next phase or Act two of Generative AI which is more focused on solving human problems end-to-end as opposed to foundational models being the “hammer” that unleashed a wave of novelty apps. It is a thoughtful follow-up on their first article on the Generative AI landscape which I have also referenced in my article “The Power of Generative AI: Exploring its Impact, Applications, Limitations, and Future”
https://www.linkedin.com/pulse/power-generative-ai-exploring-its-impact-applications-jacques-ludik/
✅ What they got right: (1) Generative AI is a thing; (2) The first killer apps emerged; (3) Developers are the key; (4) The form factor is evolving; (5) Copyright and ethics and existential dread.
❌ What they got wrong (from their assessment): (1) Things happened quickly; (2) The bottleneck is on the supply side; (3) Vertical separation has not happened yet; (4) Cutthroat competitive environment and swiftness of the incumbent response; (5) The moats are in the customers, not the data.
With respect to the latter, “workflows and user networks seem to be creating more durable sources of competitive advantage.”
Generative AI’s biggest problem is not finding use cases or demand or distribution, it is proving value.
📌 The Model Development Stack:
(1) Emerging reasoning techniques like chain-of-thought, tree-of-thought and reflexion are improving models’ ability
(2) Transfer learning techniques like RLHF and fine-tuning are becoming more accessible
(3) Retrieval-augmented generation is bringing in context
(4) New developer tools and application frameworks are giving companies reusable building blocks
(5) AI-first infrastructure companies provides plentiful GPUs at a reasonable cost, etc.
🔵 Emerging Product Blueprints:
(1) Generative interfaces
(2) New editing experiences: from Copilot to Director’s Mode
(3) Increasingly sophisticated agentic systems
(4) System-wide optimization
Democratizing Artificial Intelligence to Benefit Everyone: Shaping a Better Future in the Smart Technology Era (jacquesludik.com)
Democratizing AI Newsletter: https://www.linkedin.com/newsletters/democratizing-ai-6906521507938258944/
The Jacques Ludik Podcast
https://www.youtube.com/c/JacquesLudik
Sequoia Capital Andreessen Horowitz Google DeepMind OpenAI Anthropic Inflection AI Mistral AI Cortex Logic Machine Intelligence Institute of Africa | MIIA Cirrus AI AI Africa Consortium Compu-Connect Education AICE Africa Adanian Labs DSNai – Data Science Nigeria Google Microsoft AI Meta AI Amazon Web Services (AWS) Apple SoftBank Group Corp. Marquee Equity ONLY webinars by Cyber Gear United Nations World Economic Forum SwissCognitive, World-Leading AI Network AI for Good OECD.AI VERCHOOL HOLDINGS Future of Life Institute (FLI) Partnership on AI Vive Teens Journey Wellness Technology Innovation Institute G42 NVIDIA