Zhu Shihui Defends an AI Built on Factory Knowledge

Zhu Shihui, a manufacturer of bottle caps, recommends starting from internal knowledge within companies to introduce AI, during the AI Engineer conference in Shanghai. His talk confronts the promises of automation with outdated equipment, available skills, and industrial budgets. The event also featured demonstrations on mobile phones and presentations on robotics.
Zhu Shihui Contrasts Margins with Promises
Margins that are too low to finance a complete transformation: this is one of the constraints highlighted by Zhu Shihui for the industrial companies he describes. He criticizes consultants for presenting automation as easy, while some machines predate digital technology. According to him, this material difficulty is compounded by the lack of in-depth digital skills among the affected employees. In his company, the average employee is in their forties.
The manufacturer also emphasizes the weight of the final implementation stage, which a technical promise alone cannot resolve. Rather than an externally imposed intervention, he advocates for a modest approach built from the knowledge that teams possess without necessarily having formalized it.
Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisissez votre rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
AutoArk Runs a Model on iPhone
AutoArk showcased a demonstration running a model with 35 billion parameters on the iPhone 16 Pro. The RAM used was limited to 2.5 GB. At DYNA, Zhong Shen presented his method for building robotic hardware designed to accommodate models updated monthly.
Chinese engineers particularly rely on DeepSeek, Kimi, GLM, MiniMax, MiMo, and Qwen. The models offered by OpenAI, Anthropic, or Google are, in fact, inaccessible in China due to the filtering system known as the Great Firewall. Chinese laboratories also share their results with their counterparts, including competitors.
Ryo Lu Wants More Engaging Tools
The Shanghai edition of AI Engineer, held in early November, also hosted a discussion on the role of those who design with AI. Ryo Lu questioned the ability to know what one wants as technical possibilities multiply. His talk particularly focused on personal value, when the amount of code produced or the software tools designed no longer suffice to demonstrate it. He expressed his desire for tools that enhance his involvement, rather than establishing that he becomes less necessary.
Ying Yang, a neuroscience PhD student who is also active as an influencer, approached artificial general intelligence from a commercial angle. Followed by over 3.4 million people on Weibo, he discussed the marketing and product design opportunities available to those preparing for its arrival.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.