Welcome to insideBIGDATA’s “Heard on the Street” round-up column! This regular feature highlights thought-leadership commentaries from members of the big data ecosystem, covering the trends of the day with compelling perspectives that offer important insights for gaining a competitive advantage in the marketplace. Submissions focusing on big data, data science, machine learning, AI, and deep learning are invited. Click HERE to view previous “Heard on the Street” round-ups.
Mitigating AI Hallucinations: Ensuring Trustworthy Chatbot Interactions
Donny White, CEO and Co-Founder of Satisfi Labs, discusses the challenges of training AI agents for consistency and predictability in customer service and management. He emphasizes the importance of using the correct technology to differentiate between various AI responses to ensure trustworthiness and accuracy, especially in policy or FAQ circumstances.
Human-in-the-Loop Training Methods Will Become Increasingly Popular
Peter Stone, PhD, Executive Director at Sony AI America, predicts that human-in-the-loop training methods will become more mature and widely used in the coming years. These methods allow computers to learn from human feedback, demonstrations, evaluations, and interventions, opening up new possibilities for artificial intelligence.
The Use of AI in the Healthcare Industry
Gary Shorter, Head of AI and Data Science at IQVIA, highlights the potential of AI to improve various aspects of the healthcare industry, including clinical trials, medical diagnoses, administrative tasks, disease prevention, and telemedicine. He also emphasizes the importance of maintaining ethical standards and data privacy in the use of AI in healthcare.
LLM Models Expanding into Multimedia
Andrew Kirkcaldy, co-founder and CEO of Content Guardian, discusses the potential benefits and risks of training AI models on multimedia inputs such as images, audio, and videos. While this expansion could lead to more nuanced outputs and creativity, there are concerns about misinformation and the proliferation of DeepFakes.
Insights about the Hype Around Low-Code Platforms
Kirim Kirimli, President at Flatiron Software, shares thoughts on the impact of low-code platforms like Microsoft Power Apps on the software developer community. While these platforms make it easier to build web applications, Kirimli expresses concerns about job security for developers in the future. However, he believes that as AI advances, specific coding tasks will be handled by AI and low-code platforms.
Insights on Fine-Tuning
Pavan Belagatti, Technology Evangelist at SingleStore, discusses the importance of fine-tuning LLMs with high-quality and diverse training data to ensure robust performance. He also highlights the ethical considerations that must be taken into account during the fine-tuning process to ensure AI safety.
Source link
