In 2025, We Are Witnessing the Rapid Evolution of Generative AI and Fundamental Business Transformation
In 2025, we are witnessing the rapid evolution of generative AI technology and a fundamental transformation of the business environment. From enhanced learning features in NotebookLM's smartphone app to OpenAI's AI-powered investment banking project "Mercury," AI technology has evolved from a simple efficiency tool into a force that fundamentally transforms human knowledge work.
Particularly noteworthy is the major labor market shift occurring as the AI agent era arrives: white-collar jobs long considered high-value are increasingly being replaced by AI, while the value of blue-collar, hands-on work is rising sharply. This article provides a comprehensive overview of the AI revolution's frontline — from the latest AI tool updates and enterprise AI adoption cases to the skills business professionals need to develop going forward.
The Dramatic Evolution of AI Tools: New Productivity Through NotebookLM, Gemini, and ChatGPT
- NotebookLM's Revolutionary Updates
- Gemini's Practical Evolution
- ChatGPT's Thought Process Optimization The Dawn of the Agent Era: Innovation in Development Environments and the Frontline of Enterprise AI Adoption The Historic Labor Market Shift: "Blue Collar Billionaires" and AI's Redefinition of Occupational Value Conclusion
Looking to optimize community management?
We have prepared materials on BASE best practices and success stories.
The Dramatic Evolution of AI Tools: New Productivity Through NotebookLM, Gemini, and ChatGPT
In the 2025 AI tools market, major companies including Google, OpenAI, Adobe, and Canva are competing to release innovative features, dramatically improving business productivity. Particularly noteworthy are the major updates to Google's NotebookLM and Gemini, and OpenAI's ChatGPT.
NotebookLM's Revolutionary Updates
NotebookLM's smartphone version has been strengthened with learning support features, enabling efficient knowledge acquisition even while commuting. Newly added test and flashcard features allow you to check the contents of complex papers and technical documents through multiple-choice quizzes, powerfully supporting business professionals' continuous learning. In the PC version as well, the Gemini model running in the background has been upgraded from 1.5 Pro to 2.0 Flash, significantly increasing the number of tokens available. This enables processing of longer documents and execution of more complex analysis tasks. An auto-save feature to be implemented soon will automatically record conversation content that previously had to be saved manually in notes.
Gemini's Practical Evolution
Gemini has also evolved remarkably — the enhanced canvas feature now enables easy export of created materials to Google Slides. This is not a mere feature addition; it means the quality of AI-generated content has improved to a level where it can be directly used in practical work. The deep research feature's integration with Google Drive and Gmail has also made advanced analysis using internal data possible. Work like automatically picking up and summarizing information about AI agents from the past three months of internal documents can now be completed in just a few minutes.
ChatGPT's Thought Process Optimization
ChatGPT is keeping pace. A newly added interruption feature makes it possible to give additional instructions mid-process even during long-running tasks. This means the direction of the thought process can be dynamically adjusted, enabling more accurate results. Also important is the strengthened response to sensitive conversations. OpenAI collaborated with more than 170 specialists to improve responses to users dealing with mental health issues. Given the reality that approximately 0.07% of its 800 million weekly users — some 600,000 people — are consulting about mental health conditions, the model has been adjusted to provide more appropriate responses.
Beyond these, Google AI Studio's "Build" feature has also undergone innovative evolution. By combining Imagen (image generation), voice conversation, and Veo (video generation), a request in plain language can produce an image processing application in just a few minutes. Particularly innovative is the "annotate" feature, which allows visually commenting on created screens and giving instructions like "I want to add ten more style options." This signals the arrival of an era where even people without programming knowledge can develop practical applications.
Competition in the integration of creative tools is also intensifying. Adobe Firefly has evolved into an "all-in-one AI creative studio," providing consistent support from ideation through production to completion. Particularly noteworthy is the feature for automatically decomposing photos into parts and editing at the layer level.
Canva, meanwhile, announced "Creative Operating System" and developed "Canva Design Model" — the world's first AI model that understands design. Ask "What would make this banner more appealing?" and it provides specific design advice. Canva is also now offering Affinity completely free, accelerating the shift away from Adobe.
HeyGen's live avatar service allows you to create your own avatar for $99 and have real-time conversations in Japanese. Additionally, X.AI's Grokipedia is attracting attention as a new encyclopedia service developed by Elon Musk in response to his criticism of Wikipedia's perceived bias.
These tools are not mere feature additions — they hold the potential to fundamentally change human intellectual work. In particular, the ability to execute complex tasks through simple language instructions means that advanced work is now possible without technical expertise, and they will undoubtedly bring a revolution to how business professionals work.
The Dawn of the Agent Era: Innovation in Development Environments and the Frontline of Enterprise AI Adoption
The concept of AI agents has evolved in 2026 from mere theory to practical business tools. The transformation in development environments in particular is dramatic, and the new direction signaled by Cursor and GitHub foreshadows the future of all white-collar work.
The arrival of Cursor 2.0 has fundamentally changed the development paradigm. The newly implemented agent mode adopts a completely different approach from conventional coding assistance tools. Previously, the main development screen was in the center with an AI chat panel on the right side. In agent mode, this principal-subordinate relationship is reversed — requests to the agent are now the primary interface, with code review in a supporting role. This represents a shift from "humans write code, AI supports" to a new development style of "AI writes code, humans review and adjust."
GitHub has announced a similar direction with a concept called "Agent HQ (Headquarters)." This is an ambitious initiative to evolve GitHub from merely a code management tool into a platform for managing agents. A new feature called Mission Control allows you to oversee multiple agents working in parallel, checking each one's progress in real time. This trend suggests the future of all white-collar workers — rather than opening PowerPoint and consulting an AI agent, you consult the AI agent and see the output on the right side.
A particularly noteworthy enterprise AI adoption case is the approach taken by Mitsui Fudosan Group. The company has established a proprietary concept called "Back to Front," using AI to automate back-end processes so that humans can focus more on customer-facing touchpoints. Specifically, they have developed a "DX Division Head Agent" that embeds the thinking of their DX Division Head in AI, which employees use as a sounding board within Teams. A "Sales Support AI" combines customer data with sales experience to present optimal approach plans for initial meetings. An "FM Support AI" for condominium management associations automatically structures meeting minutes and general meeting materials, storing them in a database to enable searching for similar cases and auto-generating general meeting materials.
Behind this kind of AI adoption is the growing importance of a new job category: Forward Deployed Engineers (FDEs). FDEs are specialists who combine technical knowledge and business understanding, embedding themselves in operations to solve company-specific challenges with AI. Major AI companies like Anthropic and OpenAI are rapidly expanding hiring for this position, with FDE-related positions increasing significantly from January to September 2025. Their role goes beyond simply introducing AI tools — they understand corporate culture and existing workflows, designing and implementing AI solutions optimized for those specific contexts.
Also, a project developing a support app for visually impaired people that emerged from the Tokyo AI Festival hackathon demonstrated a new development style called "vibe coding" on smartphones. A team of five discussed and iterated in the field, repeatedly deploying and verifying on the spot to complete an app for guiding users within train stations. Starting with Google Firebase Studio and transitioning to GitHub to leverage OpenAI's Codex as the project grew more complex — this flexible development approach has the potential to become a new model for future development processes.
Large-scale AI adoption by Japanese companies is also accelerating. "SB OAI Japan," established by SoftBank Group and OpenAI, is deploying a comprehensive AI solution called "Crystal Intelligence" for large Japanese enterprises. SoftBank Group — its first customer — is advancing full-scale AI transformation with an annual investment of 400 to 500 billion yen, already experimentally operating 2.5 million custom GPTs. The vision proposed by Masayoshi Son of "1,000 AI agents per person" — like Kannon with a thousand hands — is no longer a pipe dream but a recognized achievable future.
Anthropic's establishment of a Japan subsidiary is also accelerating this trend. Its projected 2028 revenue is expected to reach $70 billion (approximately 10 trillion yen), with reports that API usage revenue has already reached twice that of OpenAI. Japanese companies including Rakuten, Mizuho Bank, Mercari, and Classmethod are adopting it in succession, and demand for enterprise AI solutions is expanding rapidly.
Furthermore, Microsoft 365 AI agent App Builder and Workflows, Google Maps' Gemini integration for enhanced navigation, the Gemini API file search tool — all major companies are competing to strengthen agent functionality. These are not mere feature additions but represent a fundamental change in how humans and AI collaborate, expected to have a major impact on overall business processes going forward.
The Historic Labor Market Shift: "Blue Collar Billionaires" and AI's Redefinition of Occupational Value
In 2025, we stand at a historic turning point in the labor market. Driven by the AI-ification of advanced knowledge work — as represented by OpenAI's secret project "Mercury" and Anthropic's "Claude for Financial Services" — the value of jobs long considered high-paying is being fundamentally reconsidered.
OpenAI's Project Mercury is an ambitious initiative aimed at AI-ifying investment banking operations. It is hiring former investment bankers and MBA holders at the premium rate of $150 per hour to create training data for advanced tasks such as financial modeling and due diligence. The true goal of this project is to replace investment banking work that currently commands annual compensation of 20 million yen with AI services costing around 5 million yen per year. Similar moves are likely to spread across a wide range of professional fields — audit and accounting, tax, legal, insurance, actuarial science, asset management, regulatory compliance, medical practice, pharmaceutical affairs, intellectual property, construction and real estate, and procurement — representing a major threat to knowledge workers.
On the other hand, the phenomenon called "Blue Collar Billionaires" is attracting attention. Cases are being reported of skilled tradespeople such as plumbers, mechanics, and electricians commanding hourly rates equivalent to lawyers and accountants — $200 to $300 per hour.
Behind this are two economic theories at work: Jevons' Paradox (productivity improvements reduce costs and increase consumption) and the Baumol Effect (the relative rise in value of fields where productivity cannot be improved). In industries where AI raises productivity, lower costs increase demand and the overall market expands. Conversely, in industries resistant to AI, relative scarcity increases and service prices rise — a structural change.
Particularly notable is that even within industries where AI adoption is advancing, complete automation is difficult — there is always a "final 10%" that only humans can do. Even if 90% of programming is AI-ified, human intervention is required for final judgment and complex problem solving, and the value of those who can handle this "bottleneck" is rising dramatically.
The most important changes in labor market value redefinition are:
- The declining value of AI-replaceable knowledge work, and the rising value of judgment and adjustment tasks within that work that AI cannot do
- The relative increase in value of physically present, hands-on work and interpersonal services that are difficult to AI-ify
- The creation and high demand for new job categories like FDE that bridge technology and business
- The scarcity value of people who can effectively utilize and manage AI agents
- The re-evaluation of distinctly human capabilities: creativity, empathy, and complex problem solving
These changes are demanding fundamental reconsideration of education systems and career development thinking. The reality of a UC Berkeley graduate getting no responses after applying to 100 internships illustrates that the conventional high-education elite track no longer guarantees success.
Conclusion
The 2025 AI revolution is transforming not merely technology but the very way we work, the way we learn, and society's fundamental values. The evolution of generative AI tools like NotebookLM, Gemini, and ChatGPT is dramatically advancing the efficiency of knowledge work, and the agent-centric development environments demonstrated by Cursor and GitHub are presenting new forms of human-AI collaboration.
In enterprise settings, movements accelerating to shift humans toward higher-value work through AI — like Mitsui Fudosan's "Back to Front" strategy — are gaining momentum. Simultaneously, in the labor market, as demonstrated by the "Blue Collar Billionaire" phenomenon, the value of skills and hands-on work that AI cannot replace is rising sharply, and the traditional occupational hierarchy is undergoing major disruption.
To survive in this turbulent era, developing the ability to collaborate with AI, maintaining a posture of continuous learning, and continuing to refine distinctly human creativity and empathy are indispensable. We now stand at a historic turning point — and what is required is not to fear this change but to embrace it as a new opportunity and actively adapt to it.
Reference: https://www.youtube.com/watch?v=avmi3q9bKRc&t=3062s
Streamline Event Management with AI | TIMEWELL Base
Struggling with the complexity of managing large-scale events?
TIMEWELL Base is an AI-powered event management platform.
Track Record
- Adventure World: Managed Dream Day with 4,272 attendees
- TechGALA 2026: Centralized management of 110 side events
Key Features
| Feature | Impact |
|---|---|
| AI Page Generation | Event page ready in 30 seconds |
| Low-Cost Payments | 4.8% fee (among the lowest in the industry) |
| Community Features | 65% of attendees stay engaged after the event |
Feel free to reach out for a consultation on streamlining your event operations.
