AI's Evolution: The Path of Transformation from 2022 to 2028
I watch the morning news program "Mōsate" on TV Tokyo almost every day. On June 30, 2025, the special segment AI Tech Geopolitics [Global Outlook] featured Mr. Obara Kazuhiro, a writer and IT critic, as a guest commentator.
In this episode, Mr. Ozawara explained three themes related to AI. One of them was "The Business Utilization of Evolving AI." After listening to his explanation on this topic, I found myself thinking a lot about it.
Inspired by Mr. Obara's discussion on current AI business utilization, this article will explore how business applications have evolved from the past to the future.
Looking Back: The AI Landscape Around 2022
AI back in 2022 was completely different from what it is today. At that time, AI was very intelligent but lacked flexibility. It was good at specific tasks, but creative work and strategic decisions still required human input.
In the past, when businesses used AI, they mainly focused on three areas:
- Automating customer support with scripted chatbots.
- Systems for detecting fraud in banking.
- Optimizing demand forecasting in supply chains.
In other words, the goal was to improve efficiency. By delegating repetitive tasks that took longer for humans to do manually to AI, companies aimed to reduce costs.
In the latter half of 2022, a major change occurred in the AI world. DALL-E and ChatGPT were made publicly available. This wasn't just the introduction of a new tool—it was the moment when AI first demonstrated "creativity." Today, we can generate text that seems like it was written by a human and images that look like they were created by professionals, simply by using text prompts.
Nevertheless, the application of AI in business at that time was still in the experimental phase. Mainly, it was used for tasks like drafting emails and summarizing long texts. The idea was to even slightly increase employee productivity. Looking back, I think we barely understood AI's true potential back then.
Screen capture from the Morning Satellite Program on TV Tokyo.
Now: A Paradigm Shift Toward Customer Value Creation
In 2024, the way we use AI has changed significantly. According to the latest survey results from the Boston Consulting Group (BCG) targeting CMOs, an interesting trend is emerging.
Recently, I've noticed that the focus of AI has completely shifted from simple cost reduction to delivering value to customers. Even while operating Kafkai, I feel this change is substantial. What customers are now seeking is not just "cheap and fast," but rather "better experiences."
According to BCG's survey, the three main areas where CMOs are focusing on AI applications are:
- Improving the quality of customer experiences (36%)
- Enhancing the quality and quantity of personalized content (33%)
- Obtaining high-quality customer insights more quickly (33%)
Particularly noteworthy is the ability of AI to read customers' emotions and needs in real time. AI can now instantly extract meaningful insights from diverse data such as social media posts, customer reviews, and support chats. This enables companies to understand almost in real time what customers are feeling and what they are seeking.
What's interesting is that this trend varies by region. While Japanese companies are still focusing on "productivity enhancement," Western companies have already started actively using AI in customer-centric applications. This difference may become a crucial factor in future competitiveness.
Looking Ahead: AI Predictions for 2028
So, what will AI look like in 2028, three years from now? Several important trends are becoming apparent.
1. The Era of Autonomous AI Agents
According to McKinsey & Company's latest report, AI is evolving from being a tool for simple tasks to an agent capable of handling complex processes. For instance, today's AI can only write an email when asked, but by 2028, an AI agent would be able to execute a new product launch campaign by conducting market research, developing strategies, creating content, distributing advertising, and measuring effectiveness—all on its own.
2. Seamless Multimodality Realization
Gartner's "Top 10 Strategic Technology Trends for 2024 mentions "Pervasive AI (Ubiquitous AI)." It is predicted that by 2028, AI will be able to process and generate various types of information—text, images, videos, and audio—smoothly and effortlessly.
For example, an AI system could analyze a customer's facial expressions during a phone call and simultaneously analyze past purchase history, providing real-time personalized recommendations. Covering everything with a single AI is incredibly convenient.
3. Expansion of Hyper-Personalization
Personalization in marketing is still in its early stages. By 2028, AI will be able to provide real-time customization of content, products, and services tailored to each individual user.
It's not just the level of "Hello, Yamada-san." AI will consider user preferences, behaviors, and even current emotions to deliver the optimal experience at the moment. For instance, suggesting content aligned with current mood or situation, or providing the best service based on behavioral patterns—such real-time responses will become the norm.
4. Rise of Specialized AI Models
Currently, there's a shift from general-purpose large language models to smaller, more efficient AI models tailored to specific industries. According to experts at the World Economic Forum, significant growth is predicted for industry-specific AI models in fields such as healthcare, law, and finance.
These specialized models will achieve higher accuracy and security. For example, in the healthcare field, AI will be able to provide highly accurate diagnostic support while protecting patient privacy, something that general AI models cannot achieve.
5. Mature Regulation and Ethical Frameworks
It is expected that comprehensive regulatory frameworks for AI development and deployment will be established by 2028. These are not just constraints but frameworks that promote the healthy development of AI. Clear rules regarding safety, privacy, and intellectual property will be introduced, making companies feel secure in their AI investments and enabling consumers to trust and use services confidently.
For example, we introduced a post on this article that discusses lawsuits regarding the intellectual property rights of data used for training large language models.
Riding the Wave of Change
Looking from 2022 to 2028, I feel that AI evolution marks the beginning of significant change. The trend of evolving from an efficiency tool to a value-creating partner and even to autonomous agents is becoming an unstoppable future.
For businesses, the key is not to fear this change but to actively embrace it. Especially for Japanese companies, the attitude of using AI to deliver value to customers is being demanded beyond just productivity improvements.
I myself am involved in AI content generation through Kafkai, and I'm amazed by the speed of this change. At the same time, this era is truly exciting. By 2028, new businesses and customer experiences that are unimaginable today will become the norm.
Are you ready? The future is just around the corner.
If you liked this article, you might also enjoy the following articles.
