Postscript of WAIC2023

After listening to the WAIC meeting for 3 days, it was an eye-opener, and it was a big lesson.

The purpose of this conference is to listen to the ideas, theories and practices of AI+ enterprise informatization. To further explore possible business scenarios. After the three-day meeting, the following points left a deep impression on me.

  • Mockups reign

The theme of the conference in 2023 has become a large model, and chatGPT is the most talked about. Metaverse, blockchain give way to big model.

  • younger

The lecturers are younger, and the group of young associate professors is very eye-catching. It shows that all universities are offering this major, but there are very few teacher resources. Or AI is new and popular.

  • Financial intelligence lags behind

In the financial intelligence forum, except for Wang Wenjing to talk about something, everyone else is far-fetched. Intelligence is a revolutionary event for financial work, and it is related to how many people's jobs. I'm ashamed of the subjects that people who have seen these speeches talk about.

 

  • Microsoft wins the AI ​​+ enterprise management software race

Microsoft stands out in this AI+ competition. Its status in the ERP field will be enhanced by people's imagination. The intelligence displayed by SAP and UFIDA is too low, which is a bit confusing. It can be regarded as a hotspot of response, but if it cannot be separated from the public model and does not consider the data security of the enterprise, it can only be a display.

  • The vertical model is the direction

Industry applications are concentrated in rich industries (medicine) and public domains (smart cities). Everyone agrees that the vertical model will be the key area of ​​innovation for SMEs in the future.

  • Domestic gap is not small

There is still a gap between domestic models. I tried a certain fire, but I was still in the "mentally retarded" stage. Tsinghua's chatGLM is better. Tsinghua revealed in the speech that it has made a trillion-parameter model, but it did not say how effective it is. It seems that the trillion-dollar war is not far away. The only applause in the entire field was given to them.

 

  • The open source community is the infrastructure to accelerate innovation

The open source community represented by Hugging Face has gathered a large number of small models. And vertical industries may need many small models. We should look for opportunities from open source models, pay attention to the hugging face, and find models suitable for specific business scenarios.

 

  • The intelligent body is the focus of the next stage

Through the mixture of large and small models, the formation of intelligent agents is a feasible path in the near future, and it is a feasible path to general intelligence.

 

  • Collision and fusion is the value of this kind of conference

If innovation comes more from the edge of the field, this conference is a hotbed of innovation. The collision of various fields, teams, and ideas will combine innovative genes and eventually hatch new species. Looking back at our educational process, what is lacking is this kind of independent learning and mutual collision mechanism.

After watching the excitement, let's go back to our original intention - the theme of AI + enterprise informatization. To summarize what we learned from it:

1. It is necessary to think about the problem from the perspective of "AI rewriting the current software".

a. Who is the software rewritten for?

When we say rewrite, do we mean the experience of giving users an "iPhone moment" or the drive that enables business owners to replace humans with AI. I think the latter is an important driver.

b. Which ones need to be rewritten?

  • Taking NetSuite as an example, the existing intelligent functions include:
    • Project Risk
    • Supply Chain Control Tower
    • AP Automation
  • Features that need to be reworked include:
    • Knowledge assistant: a knowledge assistant that can help novices get started quickly, and a personal assistant that can improve work efficiency.
    • No keyboard: user interface. No menus, let alone commands.
    • Fool-proof: When there is an abnormal input, it can be effectively reminded.
    • Interface automation: Any related application can implement code-free interface automation.
    • Insights: real-time feedback, predictions, and reminders based on data and business paradigms.
    • Process automation: a smart central scheduling system responsible for the planning and scheduling functions of the upstream and downstream of the supply chain and internal processes.

c. Rewrite process

In view of the increase in the concentration of management software, the process of AI rewriting should be a hybrid model of "core platform AI + ecological plug-ins".

  • The AIization of the core platform means that traditional major manufacturers will carry out AI transformation on their own applications. Consider what Microsoft 365 is doing.
  • The ecological plug-in refers to the new AI company using "smart agent" as the user interface to enhance the traditional products with AI. Think about how RPA does a non-intrusive interface. That's about it.

2. Need to think about the form of enterprises in the AI+ era

        Enterprise-oriented, enterprise intelligence is the end. We need to rethink the changes in the form of enterprises.

  • The biggest change is the corporate workforce. The reduction of front-line repetitive labor will lead to the reduction of background personnel. Jobs with complex skills and physical fitness will be preserved and enhanced by AI.
  • Businesses will enhance innovation outside of the business paradigm to create a unique "product or service". Don't forget that in addition to AI, there are also huge technological dividends brought about by new energy, quantum computing, new materials, genes, and brain science. Enterprise innovation is driven by these technological changes.
  • The scale of enterprises will develop in polarization due to different endowments. Large enterprises invest in the platform and rely on capital barriers to build a basic ecology, while small and micro enterprises use their strong perception ability and speed of action to innovate.

3. What should we do in the AI+ era

Don't talk nonsense, study!

 

 If you have any questions about NetSuite, welcome to talk. My email: [email protected]

Guess you like

Origin blog.csdn.net/remottshanghai/article/details/131625392