|
|
|
The 4th Industrial Revolution and History's Lessons
For this week's newsletter, I stepped back to look at the AI industry as a whole. I'm sharing the developments that continue to reinforce my strong belief that, like the printing press, the mechanical loom, the steam engine, and the computer, AI and LLMs will be major disruptive agents. But history shows that these disruptions reshaped industries creating new roles that we couldn't have imagined before. AI won't decimate our workforce -- the key is continuing to identify where our value is most needed as the capabilities, and limitations, of AI continues to be clarified.
We're getting clearer data on what AI actually does well, where it falls short, and what that means for the human-AI collaboration. The creativity research reveals AI's strength in generating volume and options at scale and the emotional intelligence data shows where human judgment and nuanced understanding remain essential. The last lead article explores how you and your teams can strengthen that collaboration, whether you're just getting started or looking to get better results from what you're already doing.
|
|
On creativity tests, ChatGPT out-scored humans nearly 3 to 1 showing AI can generate original ideas at scale, but lacks the contextual judgment to know which ideas actually work in your specific situation
|
|
AI is predicted to automate 40% of tasks by 2030, making emotional intelligence (human skills like empathy, collaboration, and strategic prioritization) even more critical
|
|
Deep Research mode and connecting AI to your email, calendars, and drives can make a big difference in collaboration quality by shifting AI from a generic assistant to a context-aware partner
|
|
|
AI Just Beat Humans at Creativity
A university study put 46 humans head to head with ChatGPT-4o, DeepSeek-V3, and Gemini 2.0 on standardized creativity tests. The results were striking. ChatGPT-4o scored 54.96 out of 57 on problem-solving tests compared to humans' average of 19.13. On creative thinking, AI's "average" ideas were more original than human ideas, and AI's "best" ideas exceeded anything humans produced.
But there's a catch: AI lacks intention, consciousness, and lived experience. It generates original combinations by recognizing patterns across massive datasets, but it doesn't understand context, feasibility, or real-world constraints. In fact, the researchers suggested that traditional creativity tests might not be appropriate for evaluating AI anymore, especially when test answers are available online.
Strategic Insight: Your value isn't in generating ideas -- AI can do that at scale. Your value is in judging which ideas actually work in your specific context, assessing feasibility, and translating those ideas into practical applications. The focus shifts to developing the judgment and contextual understanding that turns AI's creative outputs into valuable outcomes.
|
Emotional Intelligence: The Skill AI Can't Automate
According to McKinsey, AI will automate 40% of job tasks by 2030. At the same time, emotional intelligence accounts for 58% of performance across all job types. People with high EQ earn $29,000 more annually than their low EQ counterparts, and 90% of top performers have high emotional intelligence. The World Economic Forum ranks emotional intelligence, resilience, and empathy as top skills leaders need to thrive, and AI doesn't have it.
Key Takeaway: As AI handles more technical work, emotional intelligence shifts from a soft skill to a core competency. If you're hiring or developing talent, build an EQ assessment into your process.
View Full Article →
Two Features to Strengthen Your AI Collaboration
If you're overwhelmed by AI tool choices, Ethan Mollick's latest AI guide offers practical direction and is worth a detailed read.
The key insight: most people are underutilizing AI by staying in "auto" mode, which defaults to weaker models, and treats AI as a generic assistant instead of a context-aware partner. Two features make the biggest difference in collaboration quality: Deep Research mode, and connecting AI to your email, calendars, and drives so it understands your specific context.
NOTE: Be very conscious of permissions and privacy if you connect an AI to your personal content.
The Practical Angle: Better AI collaboration comes from understanding what's available and using it effectively. These aren't "advanced" techniques, they're options which influence AI's value as your partner.
Read Full Guide →
|
Quick Hits
|
AI Catches "Brain Rot" from Social Media Data
University research shows LLMs trained on viral social media content suffer measurable cognitive decay, with reasoning accuracy dropping from 74.9% to 57.2%. Worse: fine-tuning these models on clean data only partially restored performance.
|
|
AI Assistants Get News Wrong 45% of the Time
A European Broadcasting Union study tested 3,000 AI responses on news queries and found 45% had significant problems: incorrect sources, fabricated citations, and factual inaccuracies.
Quick Takeaway: When AI summarizes current events, verify the information as it is often confidently wrong.
|
|
Can You Tell if a Video Is Real or Sora AI?
OpenAI's Sora 2 makes realistic deepfakes trivially easy to create, so how can you tell if a video is real or not? You might think to look for Sora 2 watermarks or review the video's metadata, but watermarks can be removed and metadata can be stripped. Your best defense: stay vigilant and check for mangled text, disappearing objects, and physics-defying motions.
|
|
|
Industry Developments
|
AI Browser Wars Intensify with Atlas and Copilot Launches
On October 21, OpenAI launched ChatGPT Atlas, its AI web browser. Two days later, Microsoft launched its version with enhanced Copilot Mode in Edge. This isn't coincidental and further supports my belief that AI browsers are going to completely redefine how we interact with information and services.
|
|
China's Alibaba Takes Aim at US AI Dominance
Alibaba just entered the smart glasses market with $660 Quark AI Glasses competing against Meta's Ray-Ban collaboration. They also launched AI Chat Assistant, a new chatbot powered by Qwen3 models within their Quark app. This signals Alibaba's aggressive push into consumer AI products.
|
|
|
|
