Why the Future of AI Must be Human Centric

This SXSW 2026 fireside chat features Dr. Rana el Kaliouby, AI scientist, entrepreneur, and investor at Blue Tulip Ventures, in conversation with journalist Bob Safian (host of the Rapid Response podcast). The session is framed around the thesis that humanity must actively shape AI's trajectory before it reshapes humanity on its own terms. Dr. el Kaliouby argues that the AI industry has made enormous progress on cognitive intelligence (IQ) but is critically lagging on emotional and social intelligence (EQ). She cites research showing that only 7% of human communication is verbal — the remaining 93% is nonverbal cues such as facial expressions, vocal intonation, and body posture — yet virtually all AI systems today are trained exclusively on what people say, not how they say it. She calls EQ in machines the next great frontier of AI, and argues that current benchmarks are dangerously IQ-centric, issuing a call to action for the industry to develop and adopt EQ benchmarks.

The session is structured as a 'Fact or Fiction' game using video clips from Dr. el Kaliouby's own Pioneers of AI podcast, addressing five major myths. On whether AI is in a bubble, she rates it mostly fiction, distinguishing between frothy venture dynamics (pre-product companies raising at billion-dollar valuations, circular capital flows between Nvidia and OpenAI) and the real, early-stage economic opportunity in applied AI. On robots taking over jobs, she partially disagrees with Vinod Khosla's prediction of a billion bipedal robots, arguing robots will absorb repetitive and dangerous tasks — like ship welding, where human labor is scarce and risky — but that this is better framed as augmentation than replacement. On AI being bad for creators, she agrees with Mark Cuban that AI democratizes content creation and increases the value of proprietary data and human originality, noting that private data sets become genuine competitive moats when foundational models all train on the same public internet. On AI outsmarting humanity, she endorses Arianna Huffington's view that machines can and should exceed human IQ, freeing humans to invest in 'intuitive intelligence' — the embodied, somatic wisdom that technology cannot replicate. On whether AI is a boys' club, she offers the sharpest answer: this one is straightforwardly fact, not myth. She notes that if women are excluded from founding, funding, and investing in AI companies, a decade from now the economic gap will be dramatically wider. Three of her four investments at Blue Tulip Ventures are in women-led companies.

Dr. el Kaliouby outlines her three-part strategy for shaping the AI future: investing in human-centric, category-defining AI companies; storytelling through the Pioneers of AI podcast to amplify voices outside the dominant AI headlines; and convening people with disparate backgrounds to catalyze unexpected collaboration. She articulates Blue Tulip's investment thesis across three verticals: health span (sensors, data, and AI advancing every dimension of healthcare), future of work (physical AI, AI co-workers, and agentic AI transforming antiquated industries), and sustainable living (AI applied to food innovation, manufacturing, climate, and energy). On the question of AI native devices, she argues that current smartphones are pre-AI devices, and that the next platform must be perceptual, conversational, empathetic, contextual, ambient, and memory-enabled — though the winning form factor remains unknown.

The conversation covers AI therapy and companionship, world models versus large language models, and the emerging practice of paying humans to walk through their homes generating training data for embodied AI. She warns that agentic AI products like OpenAI's Operator have launched without adequate security frameworks, and that AI companions and therapy tools lack the safety guardrails necessary to prevent harm — citing tragic cases of young people being harmed through unguarded AI relationships. Her closing call to action is threefold: lean in and experiment with AI tools; demand transparency about how models are built and validated; and insist on benchmarks for environmental impact, safety, and emotional intelligence.