pplpod

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

In 2004 the Motorola Razr V3 landed among chunky plastic handsets like something dropped from orbit: impossibly thin, clad in anodized aluminum, with a keypad etched from a single sheet of metal and a snap that made hanging up feel like a statement. It turned the phone from a tool into a status symbol people left on restaurant tables. Motorola responded by flooding the market with variations through 2005 and 2006, then perfected the clamshell with the Razr 2 in 2007, just as touchscreens and the iPhone changed what a phone was for.

This episode traces the full arc of the brand. Motorola slashed prices, sold tens of millions of cheap units, starved its own research budget, and by 2011 saw its cellular division split off as Motorola Mobility. The name survived on Kevlar-backed Droid Razr slabs until flexible displays made a true folding Razr possible in 2020. From water drop hinges and split batteries to wood-backed models and a book-style Razr Fold in 2026, it is a story of hubris, survival, and design nostalgia.

• Motorola treated the V3 like a hit song it kept remixing, releasing pink, black, and Miami Ink editions instead of inventing a successor.
• The Razr 2 was two millimeters thinner, added an external touch-sensitive screen, Crystal Talk noise cancellation, and Ferrari editions, yet sales cratered against the original.
• The Motorola Razr i ran on an Intel Atom chip, a hot, power-hungry choice that struggled with Android apps written for ARM.
• The foldable Razr loops its screen into a teardrop recess inside the hinge and splits its battery into two cells, one in each half of the phone.
• LTPO displays that drop to one frame per second made a 4.0-inch always-on cover screen practical on the Razr 50 series in 2024.

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

In 2006 more than a hundred engineers, executives, and designers gathered to hold a funeral for a discontinued gadget. The gadget was Aibo, Sony's robot dog, born in the company's Computer Science Laboratory in the 1990s. Dr. Toshitada Doi and AI expert Masahiro Fujita began with a prototype called Mutant that tracked a yellow ball and threw karate strikes. Fujita's insight was that a robot built for entertainment could fail and still be loved, because owners would read its mistakes as personality.

This episode tells Aibo's life as a biography. The first model, the ERS-110, launched in May 1999 at $2,500, and all 5,000 units sold online in 20 minutes. Owners raised their dogs from a wobbly puppy stage, hackers taught them to dance, and universities fielded them in RoboCup soccer. Then new chief executive Howard Stringer cut the project, repairs ended, and aging robots began to fail. The story closes with the 2018 return of a cloud-connected Aibo that needs a monthly subscription.

• The Life software made each Aibo start as a helpless puppy and mature according to how its owner praised or scolded it through touch sensors.
• Composer Nobukazu Takemura blended animal noises with synth sounds so Aibo would seem like its own species instead of a fake dog.
• In October 2001 Sony sent a DMCA cease and desist notice to a user known as AiboPet over custom dance code, then backed down and released programming kits.
• After Sony stopped repairs in 2014, owners watched worn neck clutches cause what they called droopy head syndrome.
• The ERS-1000 has OLED eyes and can recognize up to 100 faces, but its processing happens on Sony's servers over LTE or Wi-Fi.

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

Winamp greeted new users with a bleating llama and a boast about whipping its backside. Behind the joke was a real fix for a 1997 problem: MP3 playback strained weak desktop processors. University of Utah students Justin Frankel and Dmitry Boldyrev wrapped a simple Windows interface around AMP, an efficient decoding engine written by Tomislav Uzelac, and released it as shareware. They asked users to mail a $10 check if they liked it, and by 1998 those checks brought in roughly $100,000 a month.

This episode follows the player from dorm project to 60 million users and into its long decline. It covers the cease and desist from Uzelac's PlayMedia Systems, the botched in-house decoder called Nitrane, the thousands of skins and the Milkdrop visualizer, and the $80 million sale to AOL in 1999. Then come the failed Winamp 3 rewrite, the founders' exit, a 2022 feature for music NFTs, and a 2024 source release that exposed other companies' code.

• Nullsoft answered the PlayMedia claim by swapping in Nitrane, a rushed decoder that ignored ISO standards and made music sound distorted, before licensing a proper one from the Fraunhofer Society.
• Winamp 3 was rebuilt on the Wasabi framework and broke compatibility with existing skins and plugins, so users went back to Winamp 2.
• Winamp 5 combined the version 2 engine with version 3 skinning, and the name was simple addition: two plus three.
• The 2024 code release on GitHub barred sharing modified versions and accidentally included proprietary code from Microsoft, Intel, Dolby, and Shoutcast before the repository was deleted.
• The planned Winamp 2027 is described as a subscription streaming service built on a white label platform from Deezer.

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

In 1978 Dan Bricklin sat in a Harvard Business School classroom watching a professor build a financial model in chalk. Every time one number changed, each figure that depended on it had to be erased and recalculated by hand. Bricklin imagined a grid where the computer did that work instantly. With Bob Frankston he formed Software Arts in Arlington, Massachusetts, and the two spent a winter in an attic writing the program in 6502 assembly language for the Apple II.

This episode follows VisiCalc from its June 1979 launch to its end in 1985. Businessmen walked into hobby shops and bought $2,000 computers just to run a disk that cost under $100, and the phrase killer app was born. Then publisher VisiCorp spread the program across every 8-bit machine, shipped a deliberately limited version for the IBM PC, and left room for former employee Mitch Kapor to launch Lotus 1-2-3 in 1983. Within months the pioneer was collapsing.

• Advisers told Bricklin that mainframe financial programs already existed, missing that his idea was instant, personal interaction.
• The design nearly failed at 16 kilobytes of memory until falling RAM prices led Apple to ship the Apple II with 48 kilobytes.
• In 1979, more than 25 percent of all Apple II computers sold were bought for the sole purpose of running VisiCalc.
• Lotus 1-2-3 used the full power of the IBM PC and copied VisiCalc's menu structure and keyboard commands so users could switch without relearning.
• Monthly sales fell from nearly 40,000 copies in January 1983 to 5,700 that December, and Lotus later bought Software Arts and discontinued VisiCalc.

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

In 1983, two teenagers in Edmonds, Washington, lay rigid and face down in the outfield during a baseball game, stayed that way for several innings, then got up and left. Scott Amy and Joel Marshall called it face dancing. The same odd idea surfaced again at a 1986 disco party in Berlin, in an unaired 1994 Tom Green prank in Ottawa, and in 1997 in Taunton, England, where schoolboys Gary Clarkson and Christian Langdon named it the lying down game.

This episode traces how that pointless act of stillness turned into planking once cameras and Facebook arrived. A 2007 fan page set off a contest for the craziest photo, three friends in Adelaide coined the name in 2008, and the hunt for attention pushed people toward riskier spots. The story runs through suspended hospital workers, a fatal balcony fall in Brisbane, a peak of pop culture saturation in 2011, and the moment a prime minister's thumbs up made the whole thing uncool.

• Seven night shift staff at Great Western Hospital in Swindon were suspended in September 2009 for playing the lying down game on duty and posting the photos.
• In May 2011, 20 year old Acton Beale died after falling from a seventh floor balcony in Brisbane while attempting a plank.
• The cast of The Office planked across the Dunder Mifflin set, Jay-Z and Kanye West mentioned it in Gotta Have It, and fighter Pat Barry planked in the octagon after a knockout.
• New Zealand Prime Minister John Key appeared giving a thumbs up behind his planking son Max, then said the fad was fine if done safely.
• Successors included owling, Batmanning, Dufnering, and teapotting, the last invented on purpose by teachers at Mortlake College in Australia.

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

IBM Watson won a million dollars on national television, tried to become a cancer doctor, drifted through a string of odd jobs, and was finally sold for parts. The idea began in 2004, when an IBM manager at a steakhouse saw diners abandon their meals to watch Ken Jennings on his 74 game winning streak. Under principal investigator David Ferrucci, the 2006 prototype answered about 15 percent of clues correctly. The finished machine ran on 90 servers with 16 terabytes of RAM and had to play completely offline.

This episode tells Watson's story like the biography of a child star. It explains the DeepQA system, in which hundreds of algorithms voted on hypotheses until a confidence score triggered the buzzer, and it revisits the 2011 match against Jennings and Brad Rutter. Then it follows the push into health care with Memorial Sloan Kettering and MD Anderson, where messy medical records defeated a system built for clean text. The verdict offered here is that the engineering was real and the marketing promised far too much.

• Producers made IBM build a mechanical finger to press a physical buzzer, yet Watson could still react in about eight milliseconds.
• Watson could not hear its opponents, so it once repeated the exact wrong answer Ken Jennings had just given.
• In a Final Jeopardy category on US cities, Watson answered Toronto with five question marks when the answer was Chicago, home of O'Hare and Midway.
• The MD Anderson pilot was discontinued after a reported $65 million investment without ever being used on active patients.
• Later assignments included personal shopping for an outdoor clothing brand, recipes like Peruvian potato poutine, and tax deductions, before the health unit was sold to a private equity firm in 2022.

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

In late 1994 David Bohnett and John Rezner launched Beverly Hills Internet, a small web host in Southern California with an unusual way of organizing the early web. Pages lived in virtual neighborhoods named for real places, such as Hollywood, Wall Street, and Silicon Valley, and each address was a street number with neighbors on either side. By mid 1995 anyone could claim two megabytes of free space and build a home page. The company called its users homesteaders and soon renamed itself GeoCities.

This episode traces the boom and the demolition. A forced floating watermark broke hand-coded pages in June 1998 and set off user protests just before an IPO that sent shares from $17 past $100. Yahoo bought the site in January 1999 for $3.57 billion in stock, then claimed rights to user content, scrapped the neighborhoods, and throttled free pages. When Yahoo announced the shutdown of the United States service in 2009, volunteer archivists raced to copy what they could before it vanished.

• By 1999 GeoCities was the third most visited site on the web, behind only AOL and Yahoo.
• The Federal Trade Commission found GeoCities had sold members' personal data to marketers after promising not to, including information collected from children.
• Yahoo's new terms of service asserted ownership of everything users posted, and the backlash forced a reversal within days.
• Free accounts were capped at roughly 4.2 megabytes of data transfer per hour, so a few visitors at once could knock a page offline.
• Archive Team released a 641 gigabyte torrent of rescued pages a year after the shutdown, while the Japanese version of the service survived until 2019.

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

Eric Migicovsky took his idea for a smartphone-connected watch through Y Combinator, earned real revenue during the program, and raised $375,000 from angel investors including Tim Draper. Venture capitalists still turned him down, wary of the tooling, inventory, and recall risks of hardware. So in April 2012 he renamed his company Pebble Technology and put the watch on Kickstarter with a $100,000 goal. It hit that mark in two hours and went on to raise more than $10.2 million from nearly 69,000 backers.

This episode follows the whole arc: the scramble to manufacture, the e-paper screen that gave Pebble a week of battery life, one million units shipped by the end of 2014, and a second campaign that raised $20.3 million. It also covers the reported $740 million offer from Citizen that Migicovsky declined, the 2016 insolvency, and the modest sale of software assets to Fitbit. Then comes the unlikely sequel, after Google released the Pebble OS code in January 2025 and its founder started again.

• Before Pebble, Migicovsky built a watch called the inPulse for BlackBerry devices under the company name Allerta.
• E-paper draws power only when the image changes, which let Pebble last a week while rivals struggled to get through a day.
• The Fitbit deal, reported at between $23 and $40 million, brought job offers for 40 percent of the staff and a pledge to keep Pebble's servers running.
• A fan group called Rebble reverse engineered the service and kept the watches working, funded in part by a $3 monthly subscription.
• The founder's new company recovered the Pebble trademark in July 2025, clashed with Rebble over the app archive, and open sourced its mobile app that December.

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

The Curta was a mechanical calculator small enough to hold in one hand, nicknamed the pepper grinder and the math grenade. Its inventor, Curt Herzstark, grew up in his family's calculating machine business in 1930s Vienna and set out to shrink a desktop machine to pocket size. His answer was to replace the usual row of ten stepped drums with a single central drum ringed by input sliders, and to handle subtraction by adding a nine's complement. He secured a key patent in 1938, the year Nazi Germany annexed Austria.

This episode follows what happened next. The family factory was forced to make gauges for the German military, and in 1943 Herzstark, the son of a Catholic mother and a Jewish father, was arrested and sent to Buchenwald. There a camp official offered him a desk and drafting tools: finish the calculator as a victory gift for Hitler and he might live. He redrew the whole machine from memory. After liberation in April 1945 he built prototypes, fled the Soviet zone, and fought his own investors to bring the Curta to market.

• Lifting the crank a fraction of an inch engaged a second set of complement teeth, so the drum subtracted while still turning in the same direction.
• By November 1945 machinists near Weimar had built three working prototypes from the drawings Herzstark made in the camp.
• Backers in Liechtenstein diluted his stock to nearly nothing, but they had left the patents in his name to avoid liability, which forced them to pay him.
• About 140,000 Curtas were made between 1948 and 1972, and rally navigators and pilots kept using them because vibration could not break them.
• Only about 3 percent came back for warranty repair, though many arrived in pieces after curious owners took them apart and could not rebuild them.

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

William Shockley, co-inventor of the transistor, persuaded financier Arnold Beckman to give him $1 million for a lab in Palo Alto, then mostly fruit orchards. He recruited eight young scientists, among them Gordon Moore, Robert Noyce, Eugene Kleiner, and Jean Hoerni, after putting each through psychological and intelligence tests. Then the dream job soured. Shockley recorded phone calls, forbade staff from sharing results, mailed their reports to Bell Labs to be checked, and demanded that everyone take a lie detector test.

This episode recounts the revolt that followed. Kleiner slipped off to New York to hunt for investors, the group asked Beckman to replace Shockley and was refused, and bankers Arthur Rock and Alfred Coyle met them at the Clift Hotel in San Francisco. Backed by a $1.38 million loan from Sherman Fairchild, all eight resigned on September 18, 1957, and founded Fairchild Semiconductor. Their company produced the planar process, then splintered into Intel, Kleiner Perkins, and dozens of spin-offs.

• Shockley abandoned the commercially promising bipolar transistor to chase a four-layer device he named the Shockley diode, which proved nearly impossible to manufacture.
• Consultants who reviewed the hiring tests warned that Noyce and Moore lacked the personality to be good managers.
• Coyle laid ten one dollar bills on the table and had everyone sign each one as their contract with one another.
• Hoerni's planar process buried the transistor flat in silicon under a protective oxide layer, which made the integrated circuit possible, yet Moore largely ignored it at first.
• Sherman Fairchild bought out the founders' shares for $3 million in 1959, turning them into salaried employees and setting off the departures that ended with Intel in 1968.