The Role of AI in Complementing the Human Workforce

Vishnu: How do you see AI agents complementing the human workforce in corporate settings, especially in daily work? Will AI replace humans, leading to a lot of people getting fired? Given your extensive experience, what are your thoughts on this?

Siddharth: There’s a ramp-up period for AI integration. While some envision replacing half the workforce with AI, it actually requires six months to a year of investment to gain productivity boosts. You can’t simply use a chat interface or model and expect immediate benefits. Companies need to focus on building AI agents and identifying areas to boost productivity and trim teams. This requires careful planning rather than jumping into immediate builds.

Implementing AI Beyond Chatbots

Vishnu: In India, I’ve noticed a lot of founders want to implement AI just to say they are using it. They often limit their understanding to chatbots and simple applications, which is just scratching the surface of AI’s potential.

Siddharth: Efficiency gains from AI can be significant, but they require initial work to get right. For example, using language models like GPT can make everyday tasks more efficient. However, challenges like handling large contexts and avoiding hallucinations require experience and careful structuring.

Specialization and Generalization in AI

Vishnu: From a fitness coaching perspective, while building Alpha Coach, we explored various AI applications. Many apps implemented AI as chatbots, but the lack of specialists in prompting the outcome was a big challenge. AI-generated training programs can be good, but you need experts to guide and prompt the AI correctly.

Siddharth: Specialists are essential in prompting and guiding AI in any field, whether it’s tech, fitness, or finance. AI is not a replacement for domain expertise. It requires collaboration between subject matter experts and tech professionals to build effective solutions.

The Future of AI and AGI

Vishnu: Recently, there have been more discussions about AGI (Artificial General Intelligence). How is AGI different from AI, and what impact will it have on the job market?

Siddharth: AGI is more broadly capable, performing multiple tasks across various domains, unlike specific AI models designed for narrow tasks. AGI can handle multiple workflows and make decisions on the go, behaving like a very intelligent being.

Vishnu: Will AGI create significant job losses?

Siddharth: Every major shift, like the Industrial Revolution, was followed by a period of depression with job losses. AI and AGI will likely follow the same pattern. Companies need time to build and integrate AI solutions effectively. Job losses in the immediate future might not be solely due to AI but other factors as well.

Addressing Security Concerns with AI

Vishnu: There are security concerns around using AI, especially in sensitive areas like banking. Where do we stand on this?

Siddharth: Security is a major concern. Recent developments have shown that proprietary information is often used in building large language models. Companies need to balance productivity gains with security risks and carefully evaluate how AI fits into their workflows.

Investing in AI Startups

Vishnu: There’s a lot of VC interest in AI-based startups. How do you evaluate if an AI startup is worth investing in?

Siddharth: Investors need to ensure startups have a solid foundation and understand the scalability and modularity of their AI solutions. Just labeling a company as AI-based isn’t enough; thorough tech due diligence is crucial.

Strategic Implementation of AI

Vishnu: Implementing AI effectively requires collaboration, careful planning, and understanding the specific needs of a business. It’s not about jumping on the AI bandwagon but finding where AI can truly add value and drive productivity.

Siddharth: Absolutely. AI has immense potential, but it’s essential to approach it with a strategic mindset and not just for the sake of it.