Nearly a year after the launch of ChatGPT, organizations are scrambling to adopt theGenerative AIto gain new competitive advantages or to prevent competitors from adopting the same technology. However, this raises the question:Is there still a place for traditional forms of AI, especially predictive models based on machine learning algorithms?
Source Note: The image is generated by AI, and the image is authorized by Midjourney
Moving into 2023, McKinsey'sup to dateThe AI Report indicates that generative AI has had its "breakout year," with one-third of the organizations surveyed indicating that they are already using it on a regular basis.The survey also shows that 40% organizations plan to increase their overall investment in AI as a result of the advances in generative AI. Interestingly, however, this has not led to a widespread uplift in other forms of AI, particularly traditional models based on machine learning algorithms.
Data from the Fortune-Deloitte CEO Survey shows that 55% of CEOs are evaluating or experimenting with generative AI, but only 39% of CEOs say they are evaluating or experimenting with predictive AI.These numbers piqued the interest of Forrester analyst Kjell Carlsson, who noted that generative AI is not the same asTraditional AIThere are substantive differences between the
The generative nature of generative AI is where the real difference lies. Many companies are utilizing generative AI to develop in-house assistants and chatbots based on their internal data, text and reports. In addition, surprisingly, pharmaceutical companies are also using generative AI in accelerating drug discovery.
However, the process of putting AI applications into production hasn't changed much compared to the rise of generative AI. All of those who have been working on scaling, integratingup to dateThe traditional capabilities required in terms of technology, achieving observability and transparency, and utilizing hybrid clouds for ease and cost-effectiveness are becoming more important in the generative AI space.
While data science platform vendors such as Domino are busy adapting their business models to target generative AI, there are other vendors that are capitalizing on the generative AI trend in greater depth. openAI and its business partner, Microsoft, are taking advantage of a first-mover advantage to gain a significant share of the emerging generative AI market.
The market success of generative AI vendors reflects another important difference between generative AI and traditional AI:Generative AI is now primarily something that is bought rather than built. This was the subject of a recent article by John Thomas, a data and analytics consultant on LinkedIn. He points out that while traditional AI models are mostly custom-developed, generative AI applications are largely built using base models developed by vendors.
There are other important differences between generative and traditional AI projects, including the fact that getting started with generative AI requires smaller upfront development costs and can be launched in a matter of days. In contrast, traditional AI requires higher upfront costs and a longer startup time.
Huge differences in technology, skills, cost and data types make for different use cases. Traditional AI is primarily used for tasks that are analytical in nature and involve predicting values or categorizing observations based on past data. In contrast, generative AI generates content and performs tasks with new capabilities that include generating and manipulating code, text, images, video, audio, and data.
As organizations rapidly transition from the experimental/evaluation phase of generative AI to the limited and full production phase, they will accumulate valuable knowledge on how to use this technology. However, as past experience with big data, machine learning, and traditional AI has shown, the road to productivity can have unexpected twists and turns, even without considering the known issues with generative AI in terms of disillusionment, privacy, and legal liability.
Despite the level of hype in the mainstream media that suggests we have achieved Artificial General Intelligence (AGI)finalgoals, but those who are deeply involved in big data,advancedThose in analytics and AI recognize that we are still a long way from achieving AGI. Additionally, the collective learning curve around generative AI is bound to be steep, considering that most organizations have less than a year of experience with generative AI.
In the meantime, generative AI will continue to attract almost all the attention, and traditional AI will pay the price. Once the sugar high around generative AI subsides and executives realize that it doesn't offer a quick and easy path to transformational success, while also bringing with it a host of new questions about accuracy, transparency, and legal liability, organizations will find a firmer footing as they do the hard, but necessary, work of integrating generative AI into their existing IT stacks and business models.