In this episode of Thrive in the Age of AI, co-hosts Andrew Turner and Stelios Kotakis dig deep into one of the most pressing questions in AI today: Are large language models really as smart as we think?
With Apple recently challenging the current capabilities of reasoning in models like GPT and Claude, this episode breaks down how tech giants are faring in the race to AI mastery—and where the real limitations lie.
🎧 Tune in below or read on for the highlights!
Key Takeaways
1. Apple Calls Out Reasoning Gaps in LLMs
Apple’s internal tests—such as the Tower of Hanoi puzzle—reveal that current models stumble when faced with multi-step logical tasks. As complexity increases, models either get stuck in loops or outright fail to solve the problem. Could we be hitting the ceiling of LLM capabilities for logic-heavy tasks?
“They’re trying to overthink. They can’t follow through.” – Stelios
2. The Right Tool for the Right Job
Andrew and Stelios agree: LLMs are incredible but not universal problem solvers. Sometimes they’re the right tool for the wrong job. What we need is a hybrid approach—pairing models with specialized tools, APIs, or classic algorithms to truly unlock value.
3. Claude vs OpenAI – Who’s Winning?
Stelios shares hands-on experience using Claude and OpenAI side-by-side. While Claude shines in user experience and communication, OpenAI seems slightly stronger in problem-solving—but neither is perfect.
“Neither gave me the right answer. I had to restructure the entire project myself.”
Community, Connection, and the Human Factor
Beyond the tech talk, the hosts explore how human-centered learning and community are becoming more critical than ever:
- WhatsApp Groups: Curated content, shared learnings, direct peer support
- In-Person Events: From AI for creatives to accessibility tools for visually impaired individuals, new events are on the horizon
- No selling, just connecting: A space for meaningful AI discussions without pressure
Upcoming Events & Initiatives
- AI for Creative Industries – Empowering artists, designers, and content creators with smart tools
- AI & Accessibility – Joint event with accessibility experts to showcase inclusive AI tools
- AI in Product Management – A special session hosted by Jess to explore how PMs can harness AI strategically
Bonus: Look out for a potential Tesseract Summer Party in London
Ethics, Governance, and Lifelong Learning
As AI scales, so does the need for guardrails. The Tesseract Academy is developing:
- A new AI Ethics course (with experts from King’s College)
- Expanded resources on AI governance and data strategy
- Ongoing efforts to instill creative thinking, emotional intelligence, and other irreplaceably human skills in every learner
“We’re not just building courses—we’re building resilient, lifelong learners.”
Get Involved with The Tesseract Academy
Whether you’re a CEO, a product manager, or just AI-curious—The Tesseract Academy is here to support your learning journey:
Join our newsletter
Request access to our curated WhatsApp communities
Explore courses in AI, data science, and Web3
Website: https://www.tesseract.academy
“Empowering professionals for the AI-powered future.”
🎙️ Stay Tuned
New episodes of Thrive in the Age of AI drop regularly. Make sure to subscribe on YouTube, follow our LinkedIn, and connect with us via the website.
➡️ Next up: Exploring AI & Product Management – Stay tuned!