There are 1.6 million therapists in the United States. There are 340 million people. The math does not work. So what fills the gap?

In 2026, more and more people are finding that answer in AI. Most people turn to it for information. But what happens when they turn to it for emotional support at two in the morning, when no therapist is available and the thoughts will not stop?

In this episode of Real World AI Talks by Just AI News, host and editor-in-chief Paulo Iñaki Palma sits down with Caitlin Stamatis, a clinical psychologist and Head of Research at Slingshot AI, the company behind Ash, an AI mental health tool built on clinical evidence rather than engagement metrics. Caitlin came into this field as a skeptic. What changed her mind, and what still keeps her up at night, makes for one of the most honest conversations about AI in mental health care we have had on the show.

In this episode we cover:

• Why 16% of US adults are now turning to AI for mental health support
  Is cost really the main driver, or is mental health stigma the bigger factor
• What separates Ash from a general purpose chatbot like ChatGPT or Claude
• Can AI replace a human therapist or is it filling a gap the system never covered
• What real clinical research on AI mental health tools actually looks like
  How Slingshot AI measures outcomes using gold standard validated tools
• What users should consider before sharing sensitive mental health data with any AI
• The risks of AI companions for younger teenagers and their social development
• Where the line should be drawn between AI mental health support and AI replacing human relationships
• What success looks like for Ash in five to ten years

Caitlin shares real data from over 100,000 users of the Ash app and gives an honest take on what AI still cannot and should not do in mental health care.

Real World AI Talks is hosted by Paulo Iñaki Palma, editor-in-chief of Just AI News, and is available on Spotify, Apple Podcasts, and YouTube. New episodes every week.

Most manufacturers are sitting on millions of dollars of surplus inventory and have no idea what to do with it. In this first episode of Real World AI Talks by Just AI News, host Paulo Iñaki Palma sits down with Luke Crihfield, Senior Director of Revenue at Amplio Supply Chain, to break down how artificial intelligence is changing the way enterprise manufacturers manage, appraise, and sell industrial surplus assets. From AI-powered cataloging apps to automated appraisal engines, this conversation covers what AI can realistically do today in supply chain operations and where human judgment still has to lead.

Luke explains how Amplio built a system that uses machine learning models to assess price ranges, liquidity, and logistics costs for thousands of SKUs at once, making surplus asset disposition faster and more transparent for enterprise clients. He also talks about the limits of AI in this space: why image recognition alone is not enough to catalog industrial machinery, and why an AI output is always a starting point for a decision rather than a final answer. If you want to understand how AI actually works in a real B2B environment, rather than in theory, this is the episode to watch.

The conversation also covers how Amplio is using AI internally to improve its own revenue operations. Luke shares how his sales team uses a network of Claude-powered skills connected to Zoom, Gmail, HubSpot, and Google Drive to reduce administrative work and spend more time building client relationships. This is a clear, practical look at what AI-native revenue teams look like in 2025 and what separates genuine AI adoption from just using a faster version of old tools.

Topics covered include: surplus inventory management, AI appraisal and cataloging tools, secondary market liquidity, B2B revenue automation, sustainability in manufacturing, the $11.1M Series A from Hitachi and Yamaha Ventures, and what the next generation of AI-powered supply chain platforms may look like.