.avif)
In a recent snackable learning session (by Vincent Verhalt), the current state of AI chatbots was discussed. It was about advanced AI models such as GPT and other technologies that currently dominate the landscape. The session highlighted the importance of a sober, pragmatic approach to integrating AI within Lemon.
Battle of the Bots: GPT vs. Kapa.ai
One of the most notable parts of the presentation was the comparison between two AI systems: GPT and Kapa.ai. Both systems received the same programming question: “How to define an endpoint?” Where GPT struggled with incorrect or hallucinatory answers, knew Kapa.ai providing the correct information, including the source citation. This difference comes from how both models work. Kapa.ai uses a technique called Retrieval Augmented Generation (RAG), where the model bases its responses on current and reliable sources, while GPT relies on a pre-trained dataset with no up-to-date knowledge.
What is Retrieval Augmented Generation (RAG)?
RAG is an AI technique where a model requests specific information from an external data source before generating a response. Instead of letting a large language model such as GPT rely solely on its pre-trained knowledge, the system looks for the most relevant pieces of information in, for example, a database or documentation source. This makes it possible to provide more accurate and context-specific responses, as demonstrated during the session. This is an important development for companies that want to use AI for domain-specific applications, such as technical support or internal knowledge bases.
Why is this relevant to you?
While AI models like GPT can deliver impressive performance, they're not always reliable. This is due to the fact that they often “hallucinate” and can generate erroneous information. A well-developed RAG solution can overcome these problems by using specific sources, leading to more reliable and verifiable results. This means that at Lemon, we are not just jumping on the AI hype, but are carefully investigating how we can use this technology responsibly.
The future of AI at Lemon
The session also provided insight into how AI within Lemon can be applied to our daily workflow. Some promising use cases were mentioned, such as generating feature files based on photos or using AI for advanced searches within our knowledge bases. However, these are not simple implementations. The technology must first evolve further, especially in terms of reliability and predictability.
Lemon's practical approach
It is clear that Lemon is closely following developments in the field of AI. We're continuing to experiment and test with AI in specific, controlled environments so that we're ready to integrate the technology as soon as it's mature enough. We see the enormous potential of AI, but want to make sure that implementing it into our workflows leads to valuable and reliable solutions for our customers.
.svg)
.svg)
