From a few years since we start, when we reflect on generating AI’s diffusion (Genai), 2024 will be regarded as a pseudo-time time-when business leaders have not hesitated to immerse their toes into unliked and unliked waters At the time, this is a wide range of experiments, optimistic and growth innovation, and the pigeons are first. In the global AI survey of McKinsey in the middle of 2024, 75 % of people predict that in the next few years, GenAI will lead to major or destructive changes in its industry.
Although we have learned about the advantages and limitations of Genai, it is important to remember that we are still in the evolutionary stage. The pilot plans can speed up quickly, and the construction is relatively cheap, but what will happen when these plans enter production within the scope of power in the CIO office? How will the functional use cases be executed in a relatively uncontrolled environment and how to avoid losing motivation before planning to show the results?
Common challenges from pilots to production
Given that Genai has the great potential of improving efficiency, reducing costs and improving decisions, C-SUITE’s task of functional business leaders is obvious-out and distorted. Business leaders began to work, played with GENAI functions and created their own pilot plans. The marketing team uses GENAI to create a highly personalized customer experience and automatically repeat the task. In terms of customer service, GENAI helps to solve the problem in real time for the power intelligent chat robot. The R & D team can analyze a large amount of data to discover new trends.
However, there are still many disconnection between all these potential and its final execution.
Once the pilot plans enter the track of the CIO office, the data will be carefully checked. So far, we are familiar with some common problems of Genai, such as model prejudice and hallucinations, and these problems have become big problems. CIO is responsible for data privacy and data governance of the entire organization, and business leaders are using data that may only be related to its specific key areas.
Three key things to consider before zooming up
There is no doubt that business leaders have made significant progress in the establishment of the Genai use case and provided impressive results for its specific functions, but the expansion of long -term impact is completely different. Before this journey, there are three considerations here:
1. As soon as possible (usually), including IT and information security teams
Functional business leaders usually develop blind people in their daily work and underestimate the expanded content required to expand the pilot plans to a wider range of organizations. However, once the pilot enters the production, business leaders need the support of IT and information security teams to carefully consider what may be wrong.
This is why I participated in the IT and information security teams from the beginning to help test pilots and solve potential problems. This will also help promote cross -functional collaboration, which is essential to propose confirmation bias that may occur in various functions and challenges.
2. Use real data as much as possible
As mentioned earlier, the problem of data -driven is one of the biggest obstacles to expand Genai. This is because the pilot plan usually depends on synthetic data, which may cause the expectations of business leaders, IT teams and the final CIO. Comprehensive data is a data generated by imitating real -world data. It is essentially a substitute for actual data, but there is no sensitive personal information.
Functional leaders will not always access real data, so some good skills to solve problems are: (1) avoid pilot plans that need to be conducted on the road; (2) formulate standards to prevent error data destruction/distortions Pilot results; (3) Use the company’s existing technology stack investment solutions to increase the possibility of future consistency.
3. Set real expectations
When GenAI launched ChatGPT at the end of 2022, Genai was the first publicity of the public, and people’s expectations for this technology to innovate the industry overnight were the highest. This kind of hype (whether good or bad) has been to a large extent. If their GENAI investment hopes to get further funds, the team is still under tremendous pressure.
The reality is that although Genai will be transformed, the company needs to spend technical time (and support) to start transformation. Genai is not a plug -in, its real value is not limited to clever chat robots or creative images. Companies that can successfully expand the Fenai plan will be those companies that first spend time to establish an innovative culture. The culture is preferred by short -term results.
We are together together
Although we have recently read how much knowledge about Genai, it is still a very new technology. The company should be vigilant about any supplier who claims to figure out all this. This arrogant restraint shrouded the judgment, accelerated the concept of semi -familiarity, and led to infrastructure issues that might bankruptcy business. On the contrary, when we entered Genai’s excitement, let us take some time to discuss how to expand this powerful technology on how to expand this powerful technology. By introducing IT teams in the early stage of this process, relying on real world data and maintaining reasonable return on investment, the company can help ensure that its Genai strategy is not only scalable, but also sustainable.
