Keep pulling the thread on Ali Ghodsi, Arvind Jain.
Ali Ghodsi believes that Artificial General Intelligence (AGI) has already been achieved.
Ali Ghodsi asserts that Large Language Models (LLMs) have become a commodity, comparable to gasoline from different gas stations where price is the main differentiator.
Arvind Jain predicts that AI will capture a significant portion of its revenue from the services industry, which he estimates is 25 times larger than the software industry.
Ali Ghodsi predicts that speech interaction will advance to the point of completely eliminating the need for keyboards.
Ali Ghodsi believes there is currently a bubble in the AI market.
Royal Bank of Canada developed AI agents with Databricks that can produce an equity research report within 15 minutes of an earnings call, a significant reduction from the industry standard of two hours.
Merck created a transformer model named TEDDY (Transformer Enabled Drug Discovery) that can predict missing genomes to understand gene regulatory networks for drug discovery.
7-Eleven is using AI agents to automate its marketing stack, including audience segmentation, creation of marketing materials, and campaign assembly.
Glean's internal efforts to build and fine-tune custom models for specific product use cases were unsuccessful, leading the company to adopt pre-built open-source or large foundation models instead.
Ali Ghodsi states that many startups, including some high-profile ones, have failed in their attempts to replace Robotic Process Automation (RPA) with generative AI models.
Ali Ghodsi states that prominent AI scientists like Rich Sutton and Yann LeCun believe the "scaling laws" approach pursued by frontier labs is not the correct path to achieving advanced AI.
Ali Ghodsi attributes to AI scientists like Yann LeCun and Rich Sutton the prediction that achieving true advanced AI is a 20-year problem.