Mark Cuban, the billionaire Shark Tank investor, is making a provocative prediction: the humanoid robotics industry will collapse within 5 to 10 years, despite billions of dollars pouring into companies like Figure AI, Tesla, and Boston Dynamics. His skepticism challenges the $12.3 billion in cumulative funding that has flowed into humanoid robotics companies since 2020, raising serious questions about whether the industry's current trajectory is sustainable. The timing of Cuban's prediction is particularly striking given recent industry momentum. Figure AI closed a $675 million Series B funding round in February 2024 at a $2.6 billion valuation, while Tesla's Optimus program has attracted $780 million in development funding. Boston Dynamics' Atlas has demonstrated increasingly sophisticated movement capabilities, and Agility Robotics' Digit robots are already deployed in Amazon fulfillment centers for basic material handling tasks. Yet beneath these headline achievements lies a troubling gap between what robots can do in controlled demonstrations and what they can actually accomplish in real-world commercial settings. This disconnect forms the core of Cuban's bearish outlook. What Technical Problems Make Cuban Skeptical of Humanoid Robots? Cuban's skepticism centers on three fundamental technical barriers that remain largely unsolved despite years of development effort: - Power Consumption: Current humanoid platforms like Figure-02 operate for just 5 to 8 hours on a single battery charge under light workloads, making them impractical for full-shift industrial deployment where workers typically labor for 8 to 12 hours. - Dexterous Manipulation: While companies demonstrate impressive zero-shot generalization in controlled environments, real-world tasks require robust perception systems that can handle variable lighting, occlusion, and material properties. Current vision-language-action models still struggle with basic tasks like cable management or handling deformable objects reliably. - Locomotion Stability: Even Boston Dynamics' latest Atlas iteration, widely considered the most advanced bipedal system, occasionally fails basic recovery maneuvers that human workers execute instinctively, particularly on uneven surfaces or during unexpected perturbations. These aren't minor engineering challenges that can be solved with incremental improvements. They represent fundamental physics and artificial intelligence problems that have resisted solution for decades. How Much Do Humanoid Robots Actually Cost Today? The economics of humanoid robotics reveal another critical problem supporting Cuban's prediction. Current production humanoids cost between $150,000 and $300,000 per unit, far above the $25,000 to $50,000 price point that most analysts consider necessary for widespread adoption. To put this in perspective, consider what a company would need to justify purchasing a humanoid robot. If a robot costs $200,000 and works for five years, that's $40,000 per year in capital costs alone, before accounting for maintenance, software updates, and charging infrastructure. A human worker earning $35,000 annually becomes economically competitive, which explains why humanoid adoption remains limited to pilot programs rather than large-scale deployment. Figure AI CEO Brett Adcock argues that manufacturing scale will drive per-unit costs below $100,000 by 2028, making warehouse and light manufacturing applications economically viable. However, this timeline remains optimistic given the industry's current production rates. How Many Humanoid Robots Are Actually Working Right Now? The commercial deployment numbers tell a sobering story. Fewer than 200 humanoid robots are deployed in commercial operations globally, with Agility Robotics' Digit accounting for the majority through Amazon fulfillment center trials. At their current production rate of 12 to 15 units per month, reaching the 10,000 plus unit scale necessary for meaningful market impact would take decades, not years. This production bottleneck exposes a critical flaw in the industry's growth narrative. Venture capital has poured $3.2 billion into humanoid startups in 2025 alone, despite limited revenue generation from actual robot deployments. Most companies remain in pre-production phases, selling development partnerships rather than manufactured units. Steps to Evaluate Humanoid Robot Investments Responsibly For corporate buyers and investors considering humanoid robotics, Cuban's skepticism reinforces the importance of measured evaluation rather than enthusiastic adoption. Here's how to approach this emerging technology: - Start with Pilot Programs: Early adopters like Amazon and Mercedes-Benz are pursuing limited trials while avoiding major operational commitments until technical maturity improves. This approach allows companies to learn without betting the business on unproven technology. - Demand Clear ROI Projections: Ask vendors for specific cost-benefit analyses showing when the robot will pay for itself through labor savings or efficiency gains. Be skeptical of projections that assume dramatic cost reductions without concrete manufacturing plans. - Monitor Technical Milestones: Track whether companies are actually solving the power consumption, manipulation, and locomotion problems rather than simply making incremental improvements to existing platforms. Real progress should be measurable and verifiable. - Assess Regulatory Readiness: OSHA guidelines for human-robot collaboration in industrial settings remain undefined for bipedal systems, potentially delaying deployment regardless of technical readiness. Understand the regulatory landscape in your industry before committing resources. What Would Need to Change for Humanoid Robots to Succeed? Cuban's prediction isn't inevitable. The humanoid robotics industry could prove him wrong if several critical factors align. Production costs must drop from current levels to under $50,000 per unit, battery life must extend to 16 plus hours for full-shift deployment, and manipulation reliability must improve significantly for widespread adoption across diverse industrial tasks. Tesla's humanoid program benefits from automotive manufacturing expertise and battery technology that could address power consumption challenges. Elon Musk projects Optimus production costs will reach $20,000 per unit at scale, though this timeline has proven optimistic for other Tesla product launches. The artificial intelligence foundation model approach pioneered by companies like Physical Intelligence and Skild AI could accelerate capability development through large-scale simulation and transfer learning, potentially overcoming the manual programming bottlenecks that have historically limited robot deployment. However, these optimistic scenarios require execution at a level the industry has not yet demonstrated. The gap between promise and delivery has widened rather than narrowed over the past five years. Why Does Cuban's Prediction Matter for the Industry? Cuban's skepticism, regardless of its accuracy, highlights the growing scrutiny facing humanoid robotics investments. The industry must demonstrate concrete progress toward commercial viability within the next 24 months to maintain current funding levels and market confidence. Venture capital is becoming more selective, and limited-runway startups will face pressure to show results or consolidate. The 5 to 10 year timeline also creates pressure for consolidation among the 40 plus humanoid startups currently competing for market position. Companies unable to demonstrate clear paths to profitability may face acquisition or shutdown scenarios as investors demand more rigorous business models and technical validation. The humanoid robotics industry stands at a critical inflection point. Billions have been invested based on the assumption that technical challenges are merely engineering problems waiting to be solved with enough capital and talent. Cuban's prediction suggests that some of these challenges may be more fundamental than the industry's optimists acknowledge. The next 24 to 36 months will determine whether his skepticism was prescient or premature.
