- Artificial Intelligence (AI) was a leading industry trend in 2023, with 152 S&P 500 companies citing “AI” during their earnings calls in the third quarter.
- Marco Argenti of Goldman Sachs predicts the emergence of hybrid AI and a shift towards more refined AI models in 2024.
- Vertical applications of AI are likely to attract increased venture capital funding, moving away from the expensive development of foundational models.
AI was a significant buzzword throughout 2023, stirring investor optimism and market excitement. The term “AI” was mentioned in a notable number of earnings calls made by S&P 500 companies. However, there is speculation as to whether the AI buzz will continue on into 2024.
Marco Argenti, the Chief Information Officer at Goldman Sachs, weighed in on this. He forecasts the rise of hybrid AI. Large foundational models, such as ChatGPT and Gemini, will work in tandem with smaller, refined models. For Argenti, this strategy combines the best of both worlds – foundational models have robust reasoning capabilities, while smaller models are more infrastructurally efficient and can be fine-tuned for data privacy.
This shift towards more refined AI can benefit a variety of industries. For instance, in financial services, AI models can be used to interpret complex legal documents like derivative contracts and translate them into a machine-readable format.
Argenti also predicts that the potential of AI will become more of a reality, moving from promises to tangible results. He believes that 2024 will see the fruition of many AI concepts, resulting in a return on investment. This will likely include improvements in developer productivity and customer support alongside streamlined operations.
Argenti asserts that regulation will play a key role, with a focus on maintaining security while encouraging innovation. He expects regulations to strike a balance that will not disadvantage US or other nations by overly restricting AI technology.
As for investment in AI, Argenti anticipates a shift in venture capital allocation from creating new foundational models to applying existing ones. There will likely be more investment in business-to-business and vertical applications attached to foundational models.