Conclusion First: “Whether AI is Smart or Not” No Longer Makes a Difference

In the past week, there have been three significant movements in the AI industry:

• Reid Hoffman and Marc Pincus Establish New AI Lab ‘Prentis’ aiming to raise $100 million.
• Midjourney Acquires Astrology App ‘Co-Star’.
• Cognition Acquires Conversational Design Startup ‘Poke’ and Integrates it into AI Assistant ‘Devin’.

At first glance, these three developments may seem unrelated, but they share a common thread: none of them are about “improving AI performance”.

Prentis automates everyday PC tasks. Co-Star brings the “experience” of astrology to AI. Poke focuses on designing the personality of how AI communicates.

In other words, the competitive axis of AI has completely shifted from “technical accuracy” to “quality of experience”.

This shift holds immense significance for small and medium-sized enterprises (SMEs). Why? In a world where competition is based on technical accuracy, capital strength speaks volumes. Only OpenAI can create GPT-5. However, in a world where competition is based on “quality of experience”, the companies that deeply understand their customers will win. This is a domain where local SMEs have a distinct advantage over large corporations.

—

Prentis — The Disruption of Cost Structures Signified by “Automation of Daily Tasks”

Reid Hoffman is well-known as the founder of LinkedIn and an early board member of OpenAI. Marc Pincus is the founder of Zynga. The next target for these two is the “automation of everyday computer tasks”.

While automation of coding is already being advanced by GitHub Copilot and Cursor, Prentis aims for the next level: not coding, but ordinary tasks—organizing emails, data entry, report creation, scheduling—jobs that “someone has to do, but don’t necessarily need to be done by a human”.

Consider how much time is spent on such tasks in the field of SMEs.

According to a survey by the Ministry of Internal Affairs and Communications, administrative staff in SMEs spend about 40-50% of their working hours on routine tasks. If an employee earning a monthly salary of 250,000 yen spends half their time on routine tasks, that amounts to 1.5 million yen in annual labor costs disappearing. For a company with five employees, that’s 7.5 million yen.

What happens if this can be replaced by an AI tool costing a few tens of thousands of yen per month? 7.5 million yen becomes 600,000 yen. The saved 6.9 million yen can be redirected to product development or customer service.

It is still unclear when Prentis’s product will be released, but the direction is clear. Tools in the same direction are already starting to hit the market, with no-code automation tools like Zapier, Make, and n8n incorporating AI, available for monthly fees of 10,000 to 30,000 yen.

There’s no need to wait. You can start experimenting with the tools available today.

—

Midjourney × Co-Star — Focusing on “What Experience to Deliver with AI” Rather Than “What to Create with AI”

Midjourney has acquired an astrology app. Many may wonder, “Why is an image-generating AI company getting into astrology?”

Co-Star is an app that combines NASA’s celestial data with AI to deliver personalized horoscopes to users daily. It has millions of downloads and boasts a passionate fan base, particularly among Generation Z.

David Holz, CEO of Midjourney, has not spoken much about this acquisition, but the intent is clear: Midjourney is evolving from an “image generation engine” to an “experience platform”.

Here lies an essential lesson for SMEs.

The technology of AI itself is becoming commoditized. For image generation, there’s Midjourney, DALL-E, and Stable Diffusion. For text generation, there’s GPT-4o, Claude, and Gemini. The performance differences are narrowing. In other words, simply “using AI” is no longer a differentiator.

What sets companies apart is “what experience they deliver to customers with that AI”.

Let’s consider an example. A local construction company uses AI to automatically generate 3D floor plans. Technically, this is already possible. But if that’s all they do, the neighboring construction company can do the same.

Now, consider this approach: the company interviews the homeowner about their family structure, hobbies, and how they spend their weekends, and then the AI proposes three floor plan options tailored to their lifestyle. Moreover, it includes seasonal sunlight simulations. The design of the proposal can also be adjusted to match the family’s atmosphere.

Even with the same technology, the experiences are entirely different. Whether customers feel “this company understands me” is where the competition lies.

And creating this feeling of “understanding” is something that SMEs, which interact directly with customers, excel at. Large corporations’ call centers cannot achieve this.

—

Cognition × Poke — The Era Where AI’s “Personality” Becomes a Competitive Advantage

Cognition is a startup known for its AI software engineer “Devin”. The company it acquired, Poke, specializes in designing the “personality” of AI’s conversational style and interactions.

This may seem subtle, but it is an incredibly important move.

Think about it. When you ask ChatGPT something, the responses tend to have a similar tone. They are polite, comprehensive, and safe. The same for everyone.

But real business is different. There’s a way to talk to regular customers, and a different way for first-time customers. Sometimes speaking in the local dialect builds more trust.

Designing AI’s “personality” means teaching AI to provide “customer service that reflects the company’s identity”.

This presents a significant opportunity for SMEs. Large corporations are often constrained by brand guidelines, making it difficult to sharpen the personality of their AI. In contrast, SMEs can decide with a single judgment from the CEO, “Our AI will have this kind of character”.

In fact, there’s a case where a local inn implemented an AI chatbot and intentionally set it to respond in a “young proprietress character”, resulting in a reservation rate from inquiries that was 2.3 times higher compared to traditional forms. The technology was a generic chatbot API, with a monthly cost of about 20,000 yen, and it took half a day to set up.

20,000 yen and half a day for a 2.3 times increase in reservation rate. This is the power of “AI’s personality”.

—

So, What Should You Do?

From these three news items, it is clear what SMEs should consider immediately.

1. Decide “What to Automate”

The “automation of daily tasks” that Prentis aims for is particularly beneficial for SMEs. Start by listing “tasks done daily that don’t require much thought” within your organization. Processing invoices, sorting emails, aggregating daily reports. Then, try incorporating tools like Zapier, Make, or even ChatGPT’s API integration. You can start for a monthly fee of 10,000 to 30,000 yen.

2. Design “What Experience to Deliver”

Simply implementing AI is not a differentiator. What matters is “how customers feel about it”. Those of us on the ground know what our customers struggle with and what delights them. Reflect that insight in the design of the AI. This becomes a weapon that large corporations cannot wield.

3. Infuse Your Company’s Identity into AI

The tone of AI responses, choice of words, and how proposals are made should reflect your company’s personality. If your company is formal, keep it formal; if it’s casual, keep it casual. If you are community-focused, using local dialects is also an option. This can be done starting today by simply rewriting ChatGPT’s system prompt. The cost is zero.

—

The “Democratization of AI” Is Over. Next Is “Personalization”.

2023 was the year of the “democratization of AI”. Anyone could use AI. 2024 will be the year of “high-performance AI” with advancements like GPT-4o, Claude 3.5, and Gemini 1.5.

And by 2025, the competitive axis of AI will have shifted to “personalization”. The automation of Prentis, the experience design of Co-Star, and the personality design of Poke are all pointing in the same direction.

“Using the same AI, how different experiences can we create?” — this is the next competition.

And this competition favors SMEs that can see the faces of their customers. While large corporations spend billions to create AI models, SMEs can create experiences that capture customers’ hearts with tools costing 20,000 yen and half a day of setup.

The era of technology is over. The era of experience has begun.

I want to ask: Does your company’s AI have “the same face for everyone”?