Top 6 AI Trends That Will Define 2026 (backed by data)
About this video
Jeff Su outlines six AI trends for 2026 drawn from McKinsey, OpenAI, Stanford and other reports. 1) Models matter less: frontier and open-weight models like DeepSeek and Llama are clustering in performance, and cost is falling — Nvidia chips use 105,000 times less energy per token than 10 years ago — so competition shifts to the app layer, distribution and integration (OpenAI mindshare, Google distribution via Search/Gmail/Android, Anthropic specialization). 2) Workflows beat agents: McKinsey finds no more than 10% of organizations in any function are scaling true agents, while 20% of enterprise use is already via workflow tools like custom GPTs and projects; examples include 60% less prep time and 50% fewer errors in pharma, 50% lower cost per call in utilities, and 50% fewer human hours in bank code migration, with McKinsey predicting nearly $3 trillion unlocked by 2030 from workflow redesign. 3) The technical divide ends: 75% of enterprise users report doing tasks they could not do before, and coding messages from non-technical staff grew 36% in 6 months. 4) Prompting gives way to context — the ‘fact gap’ about your goals and files — making file organization and consolidation essential. 5) Ads are coming to ChatGPT in 2026, likely as separate display banners. 6) AI moves into the physical world: Waymo has logged over 100 million fully autonomous miles with 96% fewer crashes than human drivers, Amazon warehouse robots cut order-to-shipping time by 78%, and China had deployed more industrial robots than the US and rest of world combined as early as 2023, though humanoid robots remain at least 15 years away per MIT professor Rodney Brooks.
Key takeaways
- Stop choosing AI by benchmark scores; choose by fit with your existing stack and workflow, since model performance and cost have converged into a commodity.
- Turn successful prompts into repeatable workflows for recurring deliverables, keeping humans on judgment calls — workflows, not fully autonomous agents, are delivering value now.
- Non-technical workers can now attempt previously outsourced technical work like dashboards, data cleaning and automations themselves.
- Organize and consolidate your files with clear names in one ecosystem so AI has the context it needs; prompting skill matters less than context quality.
- Physical AI is advancing in vehicles, warehouses and factories, turning capital assets into software-updated platforms, while humanoid robots remain long-term hype.
Episode details: published January 6, 2026 · 13:13 long. Summary written by Self Growth Videos from the episode itself. Business results and any income figures mentioned are Jeff Su’s own claims, shared for education — they are not typical or guaranteed.
← Previous: Master Gemini 3.1 for Work in 12 Minutes (2026) · Next: The Only AI Tools You Need (12-Minute Guide) →
All Jeff Su videos · Jeff Su profile · Jeff Su (official site)
