The hottest topic in the tech world has been reignited again. Last year, when Amazon recruited the founding team of AI startup Adept in an unprecedented way, it shocked the entire industry. This new transaction model, known as "reverse talent acquisition," allows large tech companies to achieve their goals without fully acquiring startups, by poaching core teams and obtaining technology licenses.

The key figure in this deal, David Luan, went from being a co-founder and CEO of Adept to leading Amazon's new AGI lab. Now, facing a wave of external criticism, Luan finally broke his silence and provided a strong defense of his choice in an interview with The Verge.

Amazon a (2)

When asked about this "reverse talent acquisition" trend, Luan's response was quite meaningful. He admitted that he hopes to be remembered in the future as an "AI research innovator rather than a transaction structure innovator." From his perspective, however, it is a completely reasonable strategic choice for tech giants like Amazon to "concentrate talent and computing resources" at this stage.

More thought-provoking is Luan's explanation for leaving the company he founded. He did not hesitate to state that he did not want to turn Adept into "a corporate enterprise that only sells small models." In his view, there are "four key research challenges on the path to AGI" that need to be solved, which is precisely his real mission.

The most shocking part was Luan's candid description of his resource needs. He openly stated that to solve these core issues, "each one requires a computing cluster worth hundreds of billions of dollars to run." Faced with such astronomical resource demands, he asked, "What other opportunities do I have to achieve this goal besides this?"

These words not only provide a strong defense for his career choices but also reveal the harsh reality of the current AI competition: in the ultimate race for general artificial intelligence, only tech giants with massive resources can truly participate in this game that will change the world.