Not all AI hallucination detectors are equal. Some tools actually work. Others just create the illusion of control. AI mistakes can lead to bad decisions, wasted resources, and costly errors. So, how do you pick the right tool? In my article, I compare three approaches—Pythia,

I’m thrilled to announce the launch of Gradient Descent – a podcast where we explore the cutting edge of AI, data science, and the future of technology. Ep.1 > youtu.be/ryNNUY3KhE4 In Episode 1, Alex Thomas and I dive into: - AI scaling laws - DeepSeek’s cost efficiency - The

New Gradient Descent Podcast Episode! Ep.2 – LLM as a Judge: Can AI Evaluate Itself? > youtu.be/fYB_RCxQ_YI This time, Alex and I dive into: • Why measuring LLMs is more complex than it seems • Strengths and weaknesses of using AI to evaluate AI • Risks of bias,

As more teams adopt LLMs, one challenge keeps coming up: how do you reliably detect hallucinations? We took a closer look at 5 popular tools — Pythia, Cleanlab, Galileo, SelfCheckGPT, and GuardRails AI — and broke down their approaches, strengths, and trade-offs. What we

Agentic AI forces us to rethink what it means for AI to act. In Ep. 3 of Gradient Descent, we look at agents that remember, reason, and learn — not just answer. From automation to AGI, the implications are huge. Listen → youtu.be/moj0PgYKXHs #AI #AgenticAI #AIethics

New Gradient Descent Podcast Episode! Ep. 4 – The Future of Prompt Engineering: From Prompts to Programs → youtu.be/ef3efmjhE3o This time, Alex and I dive into: • Why prompt engineering is shifting from ad-hoc strings to structured code • How #DSPy turns LLM prompts into

New Gradient Descent Podcast Episode! Ep. 5 – Fine-Tuning LLMs: RLHF vs DPO and beyond → youtube.com/watch?v=q_ZALZ… In this episode, Alex and I unpack: • Why RLHF became the standard for tuning LLM outputs • How DPO simplifies tuning by removing reward models • Why supervised

The entire Wisecube team is proud to join forces with John Snow Labs to advance the future of AI in health technology. globenewswire.com/news-release/2…

Expanding capabilities in healthcare AI - Wisecube is now part of JohnSnowLabs. Together, we’re bringing safe and effective AI solutions to market at scale — improving research, clinical decisions & patient outcomes. sliceofhealthcare.com/p/john-snow-la… #AI #HealthcareAI

🎙️ Ep.6 - A History of NLP and Wisecube’s AI Journey - is now live > youtu.be/y-tzgNv81kQ Alex and I reflect on Wisecube’s 8-year journey and 15+ years in AI - from early TF-IDF search systems and medical coding to building knowledge graphs, drug discovery, and tackling LLM

JohnSnowLabs acquires Wisecube boosting #healthcareAI with a billion-scale biomedical knowledge graph. Enhances literature review, accelerates #drugdiscovery, and improves hallucination detection for safer, accurate medical AI solutions. More: johnsnowlabs.com/john-snow-labs… #biotech

We are back with the new episode that covers First Steps with Model Context Protocol #MCP. Tune in now: youtu.be/hs1lQVu4fmI?si… #Anthropic