Mark cuban bets humanoid robots will flop within five years: 'the world isn't built for them'

Mark Cuban just dropped a grenade into the techno-optimist echo chamber: every shiny humanoid prototype you’ve seen this month—Tesla’s Optimus, China’s Agibot, Hyundai’s Atlas-killer—is heading for a five-year cliff, then oblivion. The billionaire’s math is brutal: 750 million existing homes, 20 million factories, entire cities carved for homo sapiens, not servo motors. Retrofitting that inventory, he told Business Insider, would cost more than the Apollo program and deliver less utility than a warehouse Roomba.

Cuban isn’t anti-robot; he’s anti-anthropomorphism. Inside Amazon’s fulfillment nodes, over one million non-humanoid units already shuffle inventory 24/7. They look like oversized skateboards, not ballet dancers, and they clear 30 % more picks per hour than any biped demo reel. The lesson: form follows function, not sci-fi fetish.

Why the boardrooms still chase legs

So why do Elon Musk and Sam Altman keep promising Rosie the Robot by 2030? Cuban’s answer is simple: storytelling sells shares. A human silhouette on stage triggers ancient limbic code—investors picture a servant, a soldier, a lover. Meanwhile, the unsexy winners—Amazon’s Kivas, DHL’s Locus bots—never trend on X because they lack a face to fantasize about.

The capital flow is staggering. In 2023 alone, $6.8 billion poured into humanoid startups, according to PitchBook. Cuban predicts most of that cash ends as scrap aluminum before the decade flips. His advice: if you must bet mechanical, back spiders. Ant-form factor robots can already inspect ducts, weld hulls, and crawl disaster zones—tasks bipeds botch at $80 k a pop.

The design trap no one mentions

Here’s the detail the press kits skip: every humanoid demonstration cheats. Ceilings are reinforced, floors are epoxy-flat, objects are QR-coded. Move the same robot into a Victorian brownstone or a Tokyo micro-apartment and the failure rate explodes. Cuban’s team ran the numbers: adapting the average American kitchen for bipedal torque would add $42 k to renovation costs. Multiply that across 130 million households and you’ve just invented the world’s most expensive vanity project.

Even China, Beijing’s current obsession, can’t subsidize physics. State media flaunts Agibot folding laundry, yet behind the curtain the bot requires 900 W continuous power—enough to run three refrigerators. Cuban’s verdict: “a party trick with a power cord.”

Meanwhile, the invisible swarm keeps growing. FedEx just doubled its autonomous forklift fleet. Maersk’s Barcelona terminal added 125 six-wheeled porters that never take a coffee break. They don’t look like us, and that’s precisely why they work.

Bottom line: the future of robotics isn’t coming on two legs. It’s already here, humming on wheels, treads, and rotors—out of sight, under budget, and allergic to photo ops. Cuban’s timeline is ruthless, but the warehouse metrics back him. Five years from now, the humanoids will be museum pieces, and the investors who chased mirages will be suing the storytellers who sold them the dream.