Rick Smith on Building Axon From Taser to AI Policing

Rick Smith spent seven years trying to sell a stun gun to ordinary Americans before he discovered the real customer was standing next to him in uniform the whole time. The device he and a retired NASA scientist built from Ace Hardware parts in a Tucson garage in 1993 (02:57) was supposed to let people like his mother defend themselves without a firearm. It flopped. Consumers kept asking him why, if the thing worked, the police weren't using it. So in 1999, with the company roughly $2 million in debt to Silicon Valley Bank and his father putting in $500,000, which was matched by Bruce Culver for a total of $1 million (08:32), Smith pivoted to law enforcement. That decision, made out of necessity rather than vision, is the hinge on which the rest of the story turns.

What follows is a case study in how companies actually survive, which is rarely in a straight line. Axon, then called Taser, went public in 2001 at a roughly $12 million valuation (09:50) and proceeded to triple in size for three straight years, becoming, by Smith's account, the best-performing stock in the world on a one and three year basis in 2004, the same year Google went public (13:39). Then came the fall: an SEC investigation into whether the company had misrepresented product safety, a short seller campaign, and, strangest of all, a personal assistant who was secretly feeding internal documents to short sellers who then leaked them to the New York Times (14:00). The stock dropped from $33 to $7 a share (14:25), customers stopped buying because cities got nervous about a company under federal investigation, and Smith walked into the SEC's San Francisco office unannounced, against his lawyers' advice, with boxes of documents, hoping to speed up an inquiry that was slowly bleeding his company to death (14:25).

The SEC eventually closed the case with no findings, which Smith describes dryly as "as good as it gets" (15:14), but the damage had metastasized into something worse: hundreds of lawsuits, from shareholders and from anyone ever injured by a police Taser. Market cap had cratered from a 2004 peak of about $2 billion to around $100 million (15:34). Most companies in that position settle to make the bleeding stop. Smith did the opposite. When his own insurer wanted to settle an early case for about $2 million, Axon refused, telling the insurer that if it paid Axon a smaller amount to sign away its liability, Axon would take the case to trial itself rather than let the insurer set a precedent that would invite a flood of copycat suits.

"If you settle this case, we will get swamped in litigation... we dig in like religious fanatics to defend our belief system." — Rick Smith (16:13)

Axon won, and kept winning. By Smith's account the company now runs with zero active litigation (16:49), a reversal built on a decade of choosing which fights to have rather than avoiding fights altogether.

The pivot nobody on the cap table wanted

The second rescue came from a friend, not a lawyer. Hadi Partovi, a college friend of Smith's who had made money in the late-1990s internet boom while Smith was still struggling to get Taser off the ground, joined Axon's board around 2007 and became, in Smith's words, a part-time chief technology officer for roughly five years (18:26). Smith had built hardware his whole career. Software was new territory, and the company nearly lost all its customer data in a data-center outage before Partovi talked him into moving to the cloud, first AWS, then Azure (18:50). Partovi also pushed for the name change from Taser to Axon, arguing that Taser had become "a sledgehammer brand" that meant electric weapon to the public and undersold everything else the company was building (20:06).

That everything else turned out to be the body camera, which Smith says now drives more of the company's roughly $30 to $40 billion valuation than the weapon the company was named after. The idea came out of a courtroom, not a product meeting. Co-defending a case where a man died in police custody, Smith watched the family present sympathetic photographs while the officer, bound by professional restraint, could only describe a violent, drug-fueled confrontation in flat, procedural language. Smith realized police needed a way to show, not just tell. Body cameras met fierce internal resistance at first; officers worried about being second-guessed by supervisors, what Smith calls fear of "Big Brother" (23:16). The adoption pattern that eventually won them over, according to a police union leader Smith quotes, was behavioral rather than rational: "No cop wants to wear a body camera until they've worn one for 90 days, and then they will not go on patrol without it" (24:38). The underlying reason is that officers are routinely subject to false allegations that a camera can instantly resolve, something Smith illustrates with a story from a ride-along in Phoenix where a suspect's threat to accuse an officer of rape evaporated the moment she was reminded the interaction was being recorded (24:58).

Adoption tipped nationally after the 2014 shooting of Michael Brown in Ferguson, Missouri, when both police unions and the family wanted a camera's account of events, for opposite reasons (25:43). Axon's bet on cloud infrastructure, made when governments still assumed anything stored off-premises was less secure, let the company absorb the sudden, enormous volume of footage. Smith says Axon now manages more video every day than YouTube by volume, with over a million police officers uploading roughly two hours of footage daily (20:23).

Drones, ethics boards, and the next decade

Axon's newest frontier is autonomous and semi-autonomous systems. The company has taken what Smith describes as the largest investment position in drone maker Skydio and secured exclusive rights to sell Skydio's drones into the public-safety market (47:33), enabling drones to reach a scene, in some cases, before a 911 dispatcher has finished logging the call (36:52). The more provocative project is a non-lethal, taser-equipped drone designed for scenarios like school shootings, a project that caused several members of Axon's AI ethics advisory board to resign in protest in 2022 over fears of normalizing armed robotics (41:31). Smith's counterargument is that the alternative, today, is sending armed officers to exchange gunfire, a method that kills bystanders and forecloses any non-lethal option once a human with a gun is in the room (42:39). He rebuilt the board, now called the Equity and Ethics Advisory Council, with more voices from communities most affected by over-policing (44:11), arguing that hearing objections early lets the company design in safeguards rather than argue about them after launch.

Smith's broader prediction is that robotic policing tools will be governed by "multi-key" approval systems requiring real-time sign-off from a supervising administrator before any force, lethal or not, is used (46:29), a structure he argues will make these systems less abusable than the status quo, in which an individual officer carries a loaded weapon out the door with no real-time oversight at all. David asked whether humanoid robots priced between $10,000 and $30,000 might reach commercial sale within 12 to 18 months (45:36), and Smith believes the deterrent effect of all this technology will come less from harsher punishment than from a higher probability of getting caught (40:55), a distinction criminology has long supported but that rarely survives contact with political debate.

The honest tension in the conversation, which Smith does not fully resolve, is between a company whose business model depends on more cameras, more sensors, and more government reach, and a public increasingly uneasy about exactly that. Smith's answer is procedural rather than philosophical: build in oversight, invite critics in early, and trust that democracies correct course when they overreach. Whether that is sufficient reassurance probably depends on which side of the camera a person expects to be standing on.