Summary of Key Points
This is a firsthand interview with a senior executive from Magga Technology, a leading player in the AI4S (using AI for scientific research) sector, conducted by Huxiu. It completely breaks away from the exaggerated narratives prevalent in the market that claim AI will revolutionize new drug development within a few years and that a trillion-dollar opportunity is about to materialize. The interview reveals the industry's reality based on the executive's practical experiences: the AI laboratory sector is not yet at a stage where it can showcase its capabilities or generate quick profits. Concepts such as fully autonomous smart factories, bipedal robots conducting experiments, and super-large models that can instantly create new drugs are all merely pipe dreams at this stage. The companies that will truly survive and reap the benefits of this trillion-dollar market are those that are willing to focus on the practical tasks of ensuring samples are delivered securely, laboratory equipment is properly operated, and various old devices from different brands are seamlessly integrated. The benefits of this trend will ultimately go to those who can actually apply AI to solve real problems for their clients.
---
Simplified Explanation of Key Points
1. Why reject a large order worth 200 million yuan to take on a small task of delivering samples in -40-degree weather? — The biggest pitfall in the industry is taking on projects that are beyond one's capabilities.
Many might think Magga is foolish for turning down such a large order, but in fact, this shows the prudence of industry veterans. The entire AI+automation sector has a problem: companies tend to accept any fancy requirements from clients, sign contracts, and receive advance payments, only to delay completion or settle for a small loss if they can't deliver. However, He Yuan sees things differently. Even the most advanced fully autonomous smart factories are not 100% reliable in the industry. Accepting such projects would damage the company's reputation and result in no profit. On the other hand, the seemingly trivial task of delivering samples in extreme weather conditions builds valuable experience—robots that can operate stably in harsh environments and adapt to different clients' needs—which can be applied to more projects. The industry is still in the learning phase, and taking on projects beyond its capabilities is like planting landmines for itself.
2. Global pharmaceutical companies are flocking to China for AI laboratories, not just for lower costs.
Many assume that Chinese AI automation companies have an advantage due to cheaper labor. However, this is a misconception. Overseas, custom AI laboratory projects are extremely inefficient: engineers may take months to schedule and weeks to make code changes, and repairs can take weeks. Chinese teams, on the other hand, respond incredibly quickly. Major pharmaceutical companies around the world have visited China and no longer focus on hype; they look for real-world examples of AI solutions and whether the company can integrate them into their processes. They care about whether the company has worked with other leading clients and can provide timely support if issues arise. What they need is a tool that can integrate into their daily research and development.
3. The trillion-dollar AI pharmaceutical sector lacks not advanced models but practical capabilities.
The industry is focused on developing fast models, but the real challenge is handling the many practical issues in the real world. For example, robots may accidentally break samples, and old experimental equipment may not communicate effectively. The most valuable aspect of an AI laboratory is its reliability and ease of use. Companies are more interested in robots that can perform tasks consistently and efficiently, even in harsh conditions.
4. Pharmaceutical companies are investing heavily in AI laboratories with clear goals in mind.
They are not just looking for cheaper labor; they want to improve efficiency and obtain high-quality, traceable data for their AI models. Projects that meet these three criteria (strategic benefits, efficiency, and data quality) have a high customer repurchase rate. Companies that can provide these benefits will have a lasting advantage in the market.
5. Don't trust financing or demo videos; judge AI4S companies by their actual capabilities.
Many companies in this sector are just trying to profit from the hype. He Yuan suggests a simple criterion: look for real paid orders from industry leaders. These orders demonstrate that the company has successfully integrated AI into real processes and can provide ongoing support. This is a true reflection of customer satisfaction, as customers are more discerning than investors. No company can monopolize the entire market; only those that can provide reliable and efficient solutions will thrive.
---
In summary, this analysis highlights the practical realities of the AI4S sector, emphasizing the importance of reliability, efficiency, and practical solutions over flashy technology.