Orient Securities: Expectations for big model makers entering the market and commercialization of the intelligent industry are expected to increase

The Zhitong Finance App learned that Orient Securities released a research report saying that GM's big model is also expected to seize market opportunities in the physical field. The bank believes that the entry of big model manufacturers will accelerate the progress of the physical model, so expectations for commercialization of the industry are continuing to rise. The bank sees that major model makers such as OpenAI are moving into the field of embodied intelligence. The bank expects this to drive breakthroughs in more application scenarios, thereby driving breakthroughs and acceleration in the commercialization process of the industry, and manufacturers of complete robot machines and components are expected to benefit.

Orient Securities's main views are as follows:

The expansion of large model robot applications has boosted the industry's commercialization expectations

The main reason the bank saw that the market was pessimistic about the embodying intelligence industry before is that it believes that the embodying model is different from the big model, and is worried that the model of embodying industry manufacturers is progressing slowly now, so application scenarios for personalized intelligence are limited. However, the bank saw that GPT-6 Astra has achieved good application practices in the physical field, and some investors are beginning to gradually believe that GM's big model is also expected to gain market opportunities in the physical field. The bank believes that the entry of large model manufacturers will accelerate the progress of the physical model, so expectations for commercialization of the industry continue to rise.

Big models have great potential as an embodying strategy layer, and emblematic manufacturers benefit the most marginally

Galaxy GM previously tested Astra and 0.5 into a two-armed RoboDojo subset with ten tasks. The results showed that the success rate of using Astra alone was only 26%. After passing the mixed control of Astra and 0.5, the task success rate increased to 48%. The bank believes that the experimental results revealed to a certain extent the potential of GPT-6-Astra as a strategic brain, that is, using Astra to understand tasks, and thereby use more basic physical models to complete mission goals. For physical body manufacturers, the underlying models of their training can be combined more efficiently, further improving deployment capabilities during actual implementation. The bank believes that along with the continuous exploration of the field of artificial intelligence by large GM model manufacturers in the future, it is expected to drive the accelerated application of artificial intelligence in logistics and industrial manufacturing, and that the barriers for manufacturers with know-how in vertical scenarios will be relatively higher.

Big model makers are speeding up the deployment of the smart field, which is expected to boost the industry's commercialization expectations

The bank sees that recently many large model makers are continuing to expand their layout in the field of physical intelligence, which is expected to catalyze an increase in the industry's commercialization expectations. OpenAI recently announced that it will launch a brilliant demonstration robot prototype in 2027. Tesla is expanding the Optimus robot acquisition team, and the domestic manufacturer LibyteDance is also expanding the robot engineer team. The bank sees that with these large model manufacturers accelerate their deployment into the field of physical intelligence, the bank expects more large model manufacturers to accelerate their deployment into the field of physical intelligence in the future. The bank believes that this is expected to drive breakthroughs in more application scenarios, thus driving the acceleration of the industry's commercialization process, and breakthroughs in robot machines and parts manufacturers are expected to benefit.

Risk Alerts

The risk of model development and data accumulation being slower than expected, the risk of unclear implementation of scenario requirements, the risk that the manufacturer's production falls short of expectations, the risk of national policy changes, and the risk of industry financing falling short of expectations