AI-podden

In this March news episode, our host Ather Gattami sits with Anders Arpteg to unpack a fast-moving period in AI, from NVIDIA GTC and the rise of agentic, always-on systems to the growing demand for inference, physical AI and autonomous driving. They argue that while the industry is moving quickly and companies are racing to build more proactive, capable systems, current models are still far weaker at reasoning and action-taking than the hype often suggests, something they link to the poor results on Arc AGI 3. The conversation also touches on shifting power dynamics between companies like Anthropic, OpenAI and Meta, before ending on a more forward-looking note with Semantic Tube Prediction, a research idea that could offer a more efficient path to reasoning in latent space rather than through token-by-token prediction.

In this February news update, our host Ather Gattami sits with Anders Arpteg, to recap the latest and most significant shifts in AI. On the menu; OpenAI's $100 billion funding round, intensifying competition from companies like Anthropic, Google, and xAI, and the growing geopolitical role of AI in defense and national security. They also explore the gap between rapid technological progress and slower real-world adoption, the evolving role of software engineers as AI automates more coding tasks, and Sweden's new AI strategy aimed at becoming a leading AI nation through stronger adoption and practical application of AI.

In this months January news update, hosts Ather Gattami and Anders Arpteg, recap the latest and most significant shifts in AI. They including new model releases, the shift toward autonomous agent systems like OpenClaw, and breakthroughs in reasoning such as Gemini DeepThink's results. They also discuss the growing concentration of AI power among tech giants, the role of open and specialized models, and why rapid adoption will be critical for companies and individuals as AI capabilities accelerate.

In this episode, Anders Hammarbäck, co-founder of Redpine AI, discusses how data, not models - will define the winners in the AI era. He explains RedPine's role as a "knowledge layer" that unlocks proprietary, high-quality data to improve AI accuracy and reduce hallucinations, especially in sectors like healthcare and pharmaceuticals. The conversation also covers the potential value of personal data, the near-term outlook for AI and AGI, Europe's position in the global AI race, and Ander's shift from venture capital to founding a company focused on data as the core competitive advantage in AI.

Happy New Year listeners! In our first AI-Podden news update of 2026, our hosts Ather Gattami and Anders Arpteg, reflect on the key highlights of AI in 2025 and what lies ahead. They discuss OpenAI's loss of momentum, Google Gemini's rise, the hype and shortcomings of agentic AI, and the growing gap between rapid technological progress and real-world adoption. The conversation also covers infrastructure and hardware investments, energy efficiency, AI sovereignty, regulation, geopolitics, robotics, and the shift toward smaller, more specialised models, with a clear warning that the AI divide between leaders and laggards is likely to widen further in 2026. It promises to be a year full of twists, turns and surprises in the world of AI.

In this months November news update, our hosts; Ather Gattami and Anders Arpteg, recap November's biggest AI shifts. Google's Gemini 3 Pro taking the lead, OpenAI facing pressure over massive infrastructure bets and AGI expectations, Yann LeCun leaving Meta for new research ventures, the EU softening its regulatory stance, and Grok sparking laughs with overly pro-Elon responses, capturing both rapid progress and rising bubble concerns.

In October's AI news update, Ather Gattami and Anders Arpteg discuss OpenAI's transition to a full for-profit structure, the rollout of in-chat shopping, and its new browser. They clarify what Google's "Quantum Echoes" actually demonstrates and why practical quantum computing is still distant. The episode also covers recent improvements in AI systems' ability to use computer interfaces and the security implications of autonomous action-taking, new details emerging from the OpenAI leadership conflict, and NVIDIA's compact DGX-Spark supercomputer. Lastly, a quick congratulations to all of this year's AI Swede of the nominees: Hanna Linderstål Anton Osika Fabian Hedin Joel Rangsjö Marcus Wallenberg Leonora Vesterbacka

*]:pointer-events-auto [content-visibility:auto] supports-[content-visibility:auto]:[contain-intrinsic-size:auto_100lvh] scroll-mt-[calc(var(--header-height)+min(200px,max(70px,20svh)))]" dir="auto" tabindex="-1" data-turn-id= "request-WEB:695faadf-e8e6-439d-a551-6a261d8a0c51-18" data-testid= "conversation-turn-4" data-scroll-anchor="true" data-turn= "assistant"> In this special episode, our host Ather Gattami interviews Jayesh Govindarajan, EVP at Salesforce AI and AgentForce, live at Dreamforce in San Francisco, where Salesforce unveiled major advances in agentic AI including AgentForce Voice, advanced context engineering, and AFScript. Jayesh explains that successful enterprise AI requires not only powerful models but also well-defined goals, reliable context, actionable capabilities, and robust guardrails that deliver both creativity and precision. He describes Salesforce's full lifecycle for building, testing, and monitoring agents, which enables continual improvement and lays the foundation for future self-learning systems that can safely expand their own capabilities.

In this episode, our host Ather Gattami, speaks with Beda Grahn, Managing Director at Kyndryl Nordics & Baltics, about how they use AI to manage and modernize the core infrastructure around the globe. From banking systems to home-care platforms, Beda explains how "Kyndryl Bridge" produces over 12 million AI-driven insights monthly, enabling preventive actions and major efficiency gains, such as cutting KYC workloads by 60 percent. Ather and Beda explore the rise of agentic AI, its challenges in enterprise adoption, and the importance of leadership in driving transformation. Beda stresses that executives must lead by example, foster reverse mentorship with young talent, and embed ethical guardrails to balance innovation with security.

*]:pointer-events-auto [content-visibility:auto] supports-[content-visibility:auto]:[contain-intrinsic-size:auto_100lvh] scroll-mt-[calc(var(--header-height)+min(200px,max(70px,20svh)))]" dir="auto" tabindex="-1" data-turn-id= "request-WEB:aef98bb1-0f42-4c9d-bc2e-a0353b005fb7-3" data-testid= "conversation-turn-8" data-scroll-anchor="true" data-turn= "assistant"> In this month's AI news update episode, our hosts Ather Gattami and Anders Arpteg discuss all the latest AI breakthroughs, from OpenAI's Sora 2 and Claude 4.5 to Gemini 2.5 and Grok 4 - and the massive infrastructure race behind them, including the $500 billion Stargate project and Elon Musk's Colossus 2 data center. They explore OpenAI's move toward productisation with its "Instant Checkout" feature, Microsoft's "Vibe Working", and Google's browser-integrated Gemini, before highlighting DeepMind's progress on the Navier–Stokes problem. The episode ends on an optimistic note: AI's power lies in augmenting, not replacing, human capability.