Antonis Ramfos, from Athens Technology Center (ATC), recently participated in the special session on AI Foundation Models and Large Language Models organized by BDVA (Big Data Value Association).
The primary objectives of this workshop, held on the 29th June, as outlined by BDVA, were to address the growing influence of Large Language Models (LLMs) and Generative AI on the market, policy, and society. Recognising the importance of these topics for BDVA members, the workshop aimed to leverage the collective resources and knowledge of the community to explore various aspects. These aspects included technical challenges, privacy concerns, ethical considerations, security implications, and the broader impact of these technologies on companies, sectors, individuals, and society. Additionally, the workshop also aimed to focus on "Data-centric AI," a topic of great interest within the community. By bringing together experts and stakeholders, BDVA intended to generate further interest and resources for the topic of Data-centric AI. Following the workshop, the BDVA Board of Directors planned to evaluate and determine how to sustain these discussions within the Association going forward.
During the session, Ramfos presented an in-depth overview of the TITAN solution—an AI-powered intelligent coach designed to combat disinformation. This solution assists users in conducting investigations by providing support for critical thinking through chat interactions and microlessons. It empowers users to make their own informed decisions about the validity of content.
Key details about the TITAN solution discussed by Ramfos include:
Open-source chatbot: TITAN utilises an open-source chatbot architecture.
Training methodology: The chatbot is trained by fine-tuning LLaMA, a Large Language Model, using user-shared conversations collected from ShareGPT.
Preliminary evaluation results: A preliminary evaluation of the TITAN solution, using GPT-4 as an evaluator, demonstrated the effectiveness of the model. Vicuna-13B, the core of the TITAN solution, achieved more than 90%* quality in comparison to other well-known models such as OpenAI's ChatGPT and Google Bard.
Performance comparison: Ramfos highlighted that Vicuna-13B outperformed other models, including LLaMA and Stanford Alpaca, in over 90%* of cases.
Training cost: The cost associated with training the TITAN solution was reported to be approximately $300.
By attending the BDVA special session, Ramfos contributed valuable insights into the TITAN solution, shedding light on its capabilities and potential in countering disinformation. His presentation provided a deeper understanding of the technology's effectiveness, comparative performance, and the resources involved in its development. The workshop served as a platform for the BDVA community to exchange knowledge, address concerns, and explore future possibilities regarding LLMs, Generative AI, and Data-centric AI.
To view the presentation go to Slideshare.