Beyond the lab: will AI fix our systems?

The Human-AI Partnership: Solving Systemic Challenges with Intelligent Solutions

At Web Summit Lisbon 2025, Mr. Neil Cairns of CGTN Europe moderated a panel on AI’s role in addressing systemic challenges. Ms. Joleen Liang (Squirrel AI Learning), Mr. Jared Goodner (Akido Labs), and Mr. Andrew Filev (Zencoder) presented their AI-driven solutions in education, healthcare, and software development, highlighting practical applications.

Ms. Joleen Liang introduced Squirrel AI Learning’s Large Adaptive Model (LAM), a personalized learning system. Unlike generic LLMs, LAM uses 20 billion data points and multimodality to tailor education, fostering critical thinking. With online platforms, tablets, and 3,000 self-study centers, Squirrel AI aims to boost student engagement and academic performance, with US expansion planned.

Mr. Jared Goodner of Akido Labs detailed their hybrid tech-healthcare model, addressing the physician shortage. With 250 US physicians, Akido Labs develops an AI physician to significantly increase patient capacity. This expands access to care, particularly for underserved “safety net populations,” ensuring more individuals receive timely medical attention and reducing health inequities.

Mr. Andrew Filev, CEO of Zencoder, showcased AI coding agents that assist professional developers. These agents understand vast codebases and perform multi-repo indexing. Zencoder plans to democratize these advanced capabilities, recognizing that limited engineering capacity means many innovative ideas remain unrealized. AI, he believes, will accelerate development and broaden software creation.

The panel stressed the importance of trust in AI. Mr. Filev advocated “trust and verify” for AI coding, highlighting the need for AI to verify its own output to improve quality. Mr. Goodner emphasized transparency in medical AI, with Akido Labs’ models providing clear diagnostic logic. Ms. Liang noted that in education, trust is earned through demonstrable student progress and enhanced confidence.

Addressing capacity and AI “hallucinations,” Mr. Goodner underscored the severe impact of physician shortages, causing long wait times and inadequate care. Mr. Filev observed that engineering capacity limits innovation, but AI can accelerate development. He asserted that rigorous quality assurance, including extensive testing and AI-driven verification, is paramount to prevent errors and ensure reliability in critical systems.

On building AI, Ms. Liang confirmed Squirrel AI’s proprietary LAM, rejecting generic LLMs for education as they hinder critical thinking. Mr. Goodner detailed Akido Labs’ extensive, curated dataset for fine-tuning medical AI, enhanced by physician feedback. Mr. Filev suggested building upon frontier models and open-source contributions is more strategic than reinventing core AI, focusing on best-in-class applications.

All panelists unanimously affirmed that AI will not eliminate human roles. Mr. Filev declared himself “team people.” Mr. Goodner emphasized medicine’s inherently personal nature, with AI enhancing human connection in care. Ms. Liang concluded that teachers must adapt by integrating AI, evolving into mentors and psychologists to foster students’ emotional and soft skills, rather than fearing job displacement.