The Collapse of 'Learn Once, Secure for Life': How the AI Era is Reshaping Work as Seen at CES 2026
AI is simultaneously rewriting the rules of 'how we work' and 'investment common sense.' This sentiment dominated the keynote sessions at CES 2026. In fact, during a live recording of the All-In Podcast (January 6, Las Vegas), Bob Sternfels, Global Managing Partner at McKinsey, and Hemant Taneja, CEO of General Catalyst, discussed the end of the era where 'learning once means you're set for life.'
1. 'AI Company Growth Speed' Shatters Investment Assumptions
Taneja succinctly summarized the rapid growth of AI companies by stating, 'The world has completely changed.' By way of comparison, he noted that while it took Stripe about 12 years to reach a valuation of approximately $100 billion, Anthropic, a company they invested in, jumped from 'about $60 billion last year to hundreds of billions this year.'
What is important here, beyond the accuracy of the figures themselves, is that top VCs recognize that the 'unit of growth speed has changed.' Taneja went as far as to say that 'a wave of trillion-dollar companies is becoming a reality' for Anthropic, OpenAI, and a few others.
1-1. The 'Winning Strategy' Shifts from 'Technology' to 'Learning Speed'
In AI, companies with a faster 'learn-improve-distribute' loop achieve economies of scale before others, regardless of product differences. That is why investors look to secure data, computing resources, and distribution (adoption channels) before looking at revenue. The fact that this discussion took place at CES is proof that 'AI is no longer about gadgets, but about industrial structure.'
2. The 'CFO vs. CIO' Dilemma Facing CEOs
Sternfels introduced a raw, real-world question. McKinsey consultants often hear this from CEOs: 'Should I listen to my CFO right now, or should I listen to my CIO?'
The CFO says, 'Wait, because the return on investment isn't visible,' while the CIO urges, 'It's crazy not to adopt it. We will be destroyed.'
This conflict shows that AI is not just an 'IT investment,' but an investment that changes the rules of competition itself (barriers to entry, cost structures, and speed). Companies that wait until they can see the ROI may find that the next thing they see is the 'ROI after losing market share.'
2-1. The Solution is Not 'Small Trials' but 'Narrowing Use Cases for Company-wide Implementation'
Many companies tend to stop at PoC (Proof of Concept). Sternfels pointed out the reality that non-tech companies are 'sitting on the fence,' unable to commit to full-scale implementation.
That is precisely why you should narrow down the 'use cases' before a company-wide rollout. The shortcut is to 'use it to the fullest' in areas where value can be measured, such as sales proposals, internal knowledge searches, and summarizing customer interactions.
3. A Realistic Answer to 'Will Young People's Jobs Disappear?'
Moderator Jason Calacanis voiced the anxiety that AI might take away entry-level tasks for new graduates and young employees: 'There are people who are scared when they look at AI.'
In response, Sternfels emphasized that even if AI handles many tasks, what ultimately determines success or failure is human judgment and creativity.
Taneja went further, stating, 'The idea of learning for 22 years and working for 40 years is broken,' and urged people to accept skilling and re-skilling as a 'lifelong job.'
3-1. A Model for 'Re-skilling' That Works in Practice
The point is not to collect certifications. Here are three recommendations:
Increase the parts you delegate to AI, and elevate yourself to a 'reviewer' role(judgment, quality assurance, ethics)
Enter the customer's environment and take hold of problem definition(AI is poor at setting problems)
Show it through deliverables: The 'chutzpah/drive/passion' that Karakanis speaks of can only be conveyed through 'results achieved'.
4. The Near Future Shown by McKinsey's 'AI Colleagues'
McKinsey's own organizational design has made this discussion concrete. Sternfels spoke of the prospect of having 'as many personalized AI agents as employees' by the end of 2026.
Furthermore, the company is proceeding with '25 squared'—a reallocation to increase client-facing work by 25% and reduce back-office work by 25%. It is a structure where value shifts toward 'interpersonal/client-facing' roles rather than simply reducing headcount.
This is also set against the backdrop of recent personnel adjustments and structural transformations in the consulting industry.
The conclusion of the AI era is simple. Learning is no longer an 'event' but an 'operation.' Companies must maintain the caution of a CFO while implementing 'full-scale, focused use cases' with the sense of urgency of a CIO. Individuals must not only use AI to increase productivity but also position themselves on the side that creates value through judgment and creativity. That was the reality heard from the stage at CES 2026.