Diving into LLMs with Namee Oberst
カートのアイテムが多すぎます
カートに追加できませんでした。
ウィッシュリストに追加できませんでした。
ほしい物リストの削除に失敗しました。
ポッドキャストのフォローに失敗しました
ポッドキャストのフォロー解除に失敗しました
-
ナレーター:
-
著者:
Namee Oberst talks about how her background in law led her to explore AI for automating legal tasks. She delves into the need to democratize AI development through open-source models to bridge the gap between technical expertise and business needs (Ai Bloks). She explains the challenges of explainable AI and ensuring trust in AI applications (e.g., preventing hallucination in AI-generated content), the future of AI integration in business workflows, including Ai Bloks' recent launch of SLIMs (small models) for structured outputs. Finally, a key takeaway for aspiring lawyers - delve into AI to solve specific problems in the legal field.
In this episode:
00:16 Introduction – I talk about Namee’s background in law and finance before transitioning into questions about AI and business applications.
01:28 Namee explains that her interest in AI stemmed from realizing the potential to automate mundane tasks in legal practice.
02:53 Namee discusses the importance of democratizing AI development and how Ai Bloks bridges the gap between technical expertise and business needs by offering open-source models.
06:48 The conversation delves into the challenges of explainable AI and how Ai Bloks ensures reliability and trust in AI applications. Namee highlights the need for consistency in information and discusses techniques used by Ai Bloks to prevent issues like hallucination in AI-generated content.
09:06 Looking ahead, Namee talks about the future of AI integration in various business workflows. She mentions the recent launch of SLIMs (small models) by Ai Bloks, which are designed to provide structured outputs for better user interaction compared to traditional chatbot interfaces.
11:58 In conclusion, Namee advises aspiring lawyers looking to make a mark in the field of AI to delve deeper into technologies and reflect on what problem they would like to solve for themselves. This could be a good starting point.