This article was handcrafted the old-fashioned way, without AI assistance except for the images, so it took quite a while to finish.
2025 has just passed, and as the Year of the Horse begins, looking back and looking ahead both show that many things have changed dramatically, while even bigger changes are still unfolding.
The slow days are gone. No more. This is an era of extreme compression, and speed matters.
Product life cycles in the life sciences industry are also being compressed. With volume-based procurement pressure in front and AIDD behind, new products will keep emerging faster and faster. The promotion window for launched products is becoming shorter, precision is the only path forward, and AI is the best accelerator.
The good news is that everything has only just begun.
In 2023, AI Was a Topic and a Toy
AI felt like the story of the wolf finally arriving. From early 2023, it became a toy for a small group of people.
At that time, social feeds were often filled with examples of AI making basic mistakes, or people teasing AI in various ways.
Most discussions focused on what large models were, what AI could be used for, what was actually right, scaling laws, and similar topics.
People paid attention to model sizes, capabilities, evaluations, and constantly refreshed leaderboards.
Countless AI startups were founded, then crushed by wave after wave of new models.
Consulting firms were already publishing endless AI reports, but the key word was still "prediction."
In 2023, MeDomino built the life sciences industry's first AI + content intelligent Q&A project. It was highly innovative at the time, but looking back now, the results were only barely acceptable. Unsurprisingly, it was not adopted at scale and was not renewed.
In 2024, AI Was PoC and a Buzzword
In 2024, AI revealed countless possibilities, but it still required positive thinking and a strong dose of optimism.
In 2024, Sora appeared out of nowhere and amazed many people. But once people actually tried it, I believe most came away disappointed, just as I did.
In 2024, RAG became a hot term and was explained again and again by countless people.
In 2024, the main focus of AI applications was still content-related work.
At the end of 2024, one internet company launched a series of AI films and invited many famous directors. The promotion was loud, but the finished works felt almost ignored.
That year, MeDomino started its first AI project for a pharmaceutical medical affairs team. The client was a global Top 10 pharma company. By 2025, the project had been rolled out across the client's organization and received strong attention from its global team.
That year, MeDomino held its first internal AI competition. Our crawling team used AI + resume structuring to quickly screen candidates, meaningfully accelerating MeDomino's recruitment process and winning first prize.
That year, MeDomino encountered the question "Is AI being used to replace employees?" for the first time in a very real way.
In 2025, AI Became a Must-Have
The story reached the 2025 Spring Festival, when DeepSeek became widely known. Our strongest feeling that year was this: if AI was not in your proposal, you could barely get a meeting with the client.
In 2025, large AI models won a gold medal at the International Mathematical Olympiad. A widely discussed article in the tech community argued that large models had entered the second half, with the core issue shifting toward better evaluation. In 2025, there was even Humanity's Last Exam (HLE). Leaderboards for large models were no longer enough. After all, is the difference between 90 and 92 points really that meaningful?
Many also said that 2025 was the first year of agents. What is an agent? It is no longer simple Q&A or content generation. It does work and solves problems.
So, that year, MeDomino began working with clients to provide AI + services for their customers: HCPs.
That year, MeDomino made up its mind to move from "MeDomino Insights" to "MeDomino in Action," upgrading from providing data and insights to delivering outcomes.
In 2025, another topic became popular: GEO, or Generative Engine Optimization. The reason was simple. The public had already begun depending on AI, and AI had become a direct conclusion tool rather than just an information source. Naturally, in 2025, MeDomino launched the industry's largest-scale, most brand-diverse pharma GEO collaboration.
In 2026, AI Is Productivity
Starting in 2026, the wolf has truly arrived. AI is productivity itself, and MeDomino has officially begun delivering results.
In the past, software and data were almost ends in themselves. If they produced insights, that already counted as high added value. As for business value, one could only hope sincerity would make it appear.
But I believe that from 2026 onward, business value should be the deliverable itself. Whether it is data or AI agents, neither is the ultimate goal. Results are. Clear output of results is the most important part.
What does precision marketing need? Precise customer coverage, effective customer engagement, personalized content, and rapid correction of large-model information.
What is the business value of content management? Rapid production of personalized content and multimodal content, with better support for content consumption across channels.
What is the business value of customer insights? High-quality sales visits, efficient coverage, faster advancement of customer beliefs, and broader customer reach.
What is the business value of GEO? Information correction and brand visibility, so your brand does not lose its voice in the AI era.
These are MeDomino's core goals for 2026.
Finally, let us return to the opening line: everything has only just begun, but speed matters.
MeDomino Lu Wenqing
The 24th day of the twelfth lunar month, Year of the Snake
