Interview, Fireside Chat
Tomer Cohen: Why LinkedIn Stories Failed; How LinkedIn's Feed Was Born; AI Startups | E1019
- AI models are projected to evolve from restructuring existing knowledge to hypothesizing new scientific discoveries regarding dark matter, dark energy, Alzheimer's, quantum mechanics, and consciousness, marking a transition described as a "whole new frontier."
- The industry is expected to shift toward bundled AI products for enterprises within the next three to five years, featuring open wrappers that combine multiple services rather than requiring individual model tailoring, with a prediction that prompts will become the primary interface between humans and machines.
- LinkedIn's performance is characterized by accelerated growth, having more than doubled its member base and more than tripled its revenue over the last six years, with the "follows" feature maintaining a 200% year-over-year growth trajectory while the platform continues to innovate its UI.
- Generative AI is anticipated to significantly reshape professional landscapes by automating drudgery tasks, allowing skills learned today to become "pretty useless" for professionals functioning in workplaces 10 to 15 years from now, with projected skill set changes of 50% within five years and 25% since the time of the statement.
- While data advantages from public data sources are diminishing due to comprehensive model training, proprietary information will remain a differentiator for fine-tuning unique applications, creating opportunities for startups to rethink job functions in sectors like healthcare, retail, and education against incumbents potentially locked to legacy thinking.
- Massive productivity gains are expected as constraints on computing power and investment are alleviated, enabling AI to support creativity, build faster and more accurate products, and potentially reduce learning processes to approximately six seconds through comparative analysis.
- Foundational models are predicted to set the baseline for all subsequent models, evolving into multiple versions until superseded by new foundations, while society adapts to intimate AI interactions that access personal digital footprints, similar to the current early stages of self-driving car development.
- Microsoft and Satya are recognized for groundbreaking AI strategy and execution, with specific optimism regarding the OpenAI integration, while incumbents like Microsoft and Adobe are expected to race to lead with AI to meet customer expectations.
- 90% of individuals are expected to lack the necessary knowledge to use AI as navigational pedals for product teams, necessitating that AI talent become a critical asset for every company as organizations navigate the shift from deterministic to non-deterministic AI systems.
- The pace of technological advancement is expected to be so rapid that the five-year horizon is considered too long for planning, with a preference for focusing on the next year, as business models and information discovery mechanisms undergo dramatic evolution.