Hours after an ICE agent shot and killed Renee Good in Minneapolis, the Trump administration was already working to shape the narrative. Official White House accounts flooded social media with clips of burning flags and clashes between locals and federal immigration agents, casting protesters—not the killing—as the story.

Among the accounts amplifying that message was Johnny MAGA, a pro-Trump X account with nearly 300,000 followers. “They’re burning the American flag right now in Minneapolis,” the anonymous account claimed, sharing a clip from the White House’s official rapid response feed. “And they really expect you to believe that ICE shot an innocent civilian.”

To its audience, Johnny MAGA looked like an independent voice, another outraged supporter in the MAGA media ecosystem. The account regularly boosts Trump’s Truth Social posts and goes to bat for the administration, attacking Democrats like California governor Gavin Newsom.

But this isn’t just a regular account. Johnny MAGA appears to actually be a White House staffer named Garrett Wade who works for the Trump administration as a rapid response manager, helping to run the very same White House account his anonymous MAGA account amplifies. A phone number associated with Wade is linked to Johnny MAGA, according to a WIRED review of publicly available records, and the connection was confirmed by a source close to the White House.

Wade and the White House did not respond to requests for comment.

The Johnny MAGA account was created in September 2021, according to its X profile. (It originally used a different handle, which referenced Wade’s birth year, according to records reviewed by WIRED.) While the account’s earliest available posts focused on NFTs, it has been a consistent pro-Trump presence since at least 2022.

The operator of the Johnny MAGA account has not disclosed an official relationship with the White House while operating the Johnny MAGA X account. Multiple media outlets, including Mother Jones, TownHall, and the New York Post, have all linked out to posts on the Johnny MAGA account seemingly as organic reflections of public sentiment on political issues.

Since Trump took office last year, the Johnny MAGA account has supported administration priorities like immigration enforcement, and allies like Turning Point USA. After Trump posted a racist AI-generated video depicting Barack and Michelle Obama as apes earlier this month, the Johnny MAGA account boosted the White House’s claim that the president didn’t watch the entire video, posting, “the most obvious tell that Trump’s Truth Social post wasn’t intentional is that he would’ve posted the entire thing if he had seen it. It’s a masterpiece.”

While the Trump administration has long cultivated a growing cast of conservative creators to spread its messaging online, a White House staffer moonlighting as an anonymous MAGA influencer would blur the already hazy line further, making it nearly impossible to distinguish between official government messaging and what appears to be organic digital support. This lack of disclosure risks undermining public trust, disinformation researchers suggest.

“People have a right to know who is trying to manipulate public opinion, and they have a right to know whether or not they’re experiencing astroturf politics,” says Samuel Woolley, a University of Pittsburgh professor who studies disinformation and media ethics. “This lack of transparency and the conflict of interest surrounding this account and the lack of disclosure all amount to a breach of public trust.”

There is very little public information online about Wade, but Federal Election Commission records link him to former White House senior communications leadership. Donations from 2023 through WinRed made by a Garrett Wade from suburban Philadelphia—where Wade graduated high school—list his employer as “tech school” in March 2023 and in December that year as Opinion Architects, a digital consultancy group. The donations also list Wade as residing in the Bucks County area of Pennsylvania, where he previously lived, according to public records. The phone number associated with Wade and with Johnny MAGA—who in the past has listed his location as Philadelphia—is also geolocated to the Bucks County area.

Google is making it very easy to feel like a music producer, whether you have the skills or not. The company has announced that ProducerAI, an AI-powered music-making platform, is joining Google Labs, bringing together sound generation, visuals, and video into one experimental creative tool.

ProducerAI is powered by a preview version of Lyria 3, Google’s newest music-generation model. You can describe what you want, and the AI helps you build it, whether that means crafting beats, shaping melodies, or experimenting with entirely new sounds.

ProducerAI first launched in July 2025 to let users collaborate with an AI agent to generate music, workshop lyrics, and remix tracks from text prompts. Until now, it relied on its own underlying models. Joining Google Labs gives it access to a much larger AI toolkit.

Inside Google’s AI-powered music studio

As a part of Google Labs, ProducerAI will pull from several of Google’s models. Lyria 3 handles music generation, while Gemini powers the conversational interface that guides users through ideas and edits.

Nano Banana will generate album art, and Veo will be used to create AI-generated music videos, turning a song idea into a complete audio and visual project. Google says it will also embed SynthID watermarks into ProducerAI outputs.

This flags AI-generated audio, images, video, and text, adding transparency as AI music becomes harder to distinguish from human work. But companies like Sony have already developed a tool to detect original songs used in AI-generated tracks.

The ProducerAI team has already worked with artists like The Chainsmokers, Lecrae, and Anjulie to shape the platform. Google positions ProducerAI as an experiment, not a replacement for musicians. However, this collaboration arrives at a time when AI-generated songs are topping Billboard charts, drawing scrutiny from artists and listeners alike.

For years, the “last mile” of digital transformation has been littered with forgotten PDFs and ignored training manuals. 

Organizations spend millions on sophisticated software like SAP or Salesforce, only for employees to struggle with basic navigation. Now, as the era of agentic AI arrives, companies face a double-edged sword: they must teach human employees to collaborate with AI, while simultaneously teaching AI agents to navigate the labyrinthine interfaces of the modern enterprise.

One idea that seems to be gaining momentum among AI-forward businesses: using screen recordings and tutorials/walkthroughs of someone performing an enterprise task — be it creating a new ticket or processing an invoice — and training AI to replicate the flow based on the screen capture. Just this week, a startup called Standard Intelligence went viral on X showing an early demo of open-ended version of this for the physical and digital world.

But the truth is, there are already players tackling this problem for the enterprise itself square-on: case-in-point, Guidde, an Israel startup born during the video-centric years of the COVID-19 pandemic, today announced an oversubscribed $50 million Series B funding round led by PSG Equity to address this exact knowledge infrastructure crisis. 

Instead of feeding an agent a static PDF manual, Guidde provides high-fidelity “Video Ground Truth”—a rich stream of data captured from real human experts as they navigate complex software.

The investment signals a shift in how the tech industry views documentation—not as a static byproduct of work, but as the critical telemetry needed to train the next generation of autonomous digital agents.

Technology: from video capture to world models

At its core, Guidde is an AI Digital Adoption Platform (ADAP). However, its technological breakthrough lies in what happens behind the scenes during a recording.

Guidde isn’t just recording pixels; it is capturing every click, scroll, and latent interaction with the HTML page—the subtle pauses, the specific scroll depths, and the corrections a human makes when a system lags. This telemetry transforms raw video into a Vision-Language-Action (VLA) training set.

Meanwhile, the platform’s Magic Redaction automatically obscures sensitive data like passwords or credit card numbers during capture, ensuring materials remain secure and HIPAA-aligned.

“Every time you click a button, you drag-and-drop, you scroll, you type, we gather the interaction… all of it, we do cleanse it—there’s no private information,” explained Guidde co-founder and CEO Yoav Einav in an exclusive interview with VentureBeat.

Under the hood, the platform captures the underlying metadata and DOM (Document Object Model) changes synchronized with the video frames. The differentiator is the telemetry hidden beneath the surface.

This rich metadata creates a “digital world model” of enterprise software. And because each enterprise uses its own unique mix of apps and processes, Guidde is creating a data moat that allows enterprise agents to reason through legacy UIs with the same spatial awareness as a human, ensuring that automation actually works in a production environment rather than just a lab demo.

For a human, it’s a tutorial. For an AI agent, it is a high-fidelity map of the interface. This allows agents to “see” and reason through complex UIs the way humans do, solving the “last mile” of automation where agents previously failed due to lack of specific enterprise and in-situ usage context.

In a sense, Guidde is building a “self-driving car” like a Waymo for computer usage.

Product: three pillars of Guidd-ance

The platform has evolved into three distinct products designed to scale with an organization’s maturity:

1. Guidde Create: The engine for subject matter experts to turn workflows into documentation in minutes.

2. Guidde Broadcast: A personalized recommendation engine—often compared to Netflix—that delivers answers inside the tools people actually use. It knows who the user is and what department they are in to surface relevant content exactly when needed.

3. Guidde Discover: The newly launched “agentic” pillar. Like Waze mapping roads by observing drivers, Discover maps software routes by tracking how employees work. It understands the workflow, creates the content, and updates it automatically when the UI changes.

Training humans how to use AI — and AI using humans

The most non-obvious aspect of Guidde’s growth is its dual-purpose mission. “We’re the only platform that trains both humans and agents,” Einav stated.

As companies roll out AI tools like Microsoft 365 Copilot or ServiceNow agents, they hit a proficiency gap. One of Guidde’s largest customers revealed they were paying over $1 million a year for a sophisticated AI tool, yet “nobody knows how to use them because they did like a 30-minute training session, and then that’s it.” Guidde closes this gap by providing “bite-sized” video tutorials in the flow of work.

Simultaneously, these videos train the AI agents themselves. Foundation models like Gemini or GPT-4 often hallucinate when tasked with specific enterprise workflows because they weren’t trained on the highly specific, internal “vanilla workflows” found in private enterprise systems. Guidde provides the “starting point,” the “metadata,” and the “x, y coordinates of the button” that an agent needs to complete an action without getting stuck.

The multimodal advantage

To maintain this level of accuracy, Guidde employs a multimodal infrastructure. The system doesn’t rely on a single model; instead, it uses a “fleet” of models that evaluate one another.

• Google Gemini: Generally used for visual tasks like analyzing PDFs or PowerPoints.

• Anthropic Claude: Leveraged for writing the storyline and narrative scripts.

• Feedback Loops: When a user edits a video, that data is fed back into the model to prevent the same mistakes from occurring in future captures.

This approach allows Guidde to replace a legacy stack of six or seven disconnected tools—Loom for capture, Adobe Premiere for editing, 11Labs for text-to-speech, and Synthesia for avatars—with a single, AI-native platform. “We basically pack everything for you,” Einav says, “and automate the entire process based on your brand guidelines.”

Video-first origin story

The genesis of Guidde lies in a frustration familiar to any product leader. Before founding the company, Einav and co-founder Dan Sahar spent years mastering video traffic at Qwilt, a company they started in 2010 to analyze how people watched Netflix and Disney+. 

When COVID-19 hit, they saw a massive opportunity to apply that video expertise to the workplace. They observed that short video explainers could increase free-to-paid account conversions by 30%, but the friction of creating them was unsustainable.

In an interview, Einav recalled the “tedious work” of the old world: “My team in Israel were creating the content, someone in the US with a US accent was doing the narration, someone in the marketing team would write the script… and someone in the enablement team would do the edit.” This fragmented workflow meant a single video took two to three weeks to produce. “And then two weeks later, the product changes, and you need to redo it from scratch,” Einav added.

Guidde was built to collapse this cycle into seconds. By automating the “Magic Capture” of a workflow, the platform generates a structured narrative script and professional AI voiceover instantly. This removes the editing bottleneck, transforming subject matter experts into “training powerhouses.”

Licensing and market impact

Guidde’s pricing structure reflects its transition from a utility to a core piece of enterprise infrastructure:

• Free: $0 (Up to 25 videos, web-app support).

• Pro: $18/creator/month (Unlimited videos, brand kits).

• Business: $39/creator/month (Unlimited text-to-voice, analytics).

• Enterprise: Custom pricing (Multi-language translation, SSO, Magic Redaction).

The platform’s impact is already visible in the numbers: a 41% reduction in video creation time and 34% fewer inbound support tickets. 

For customers like Emerson, this translates to 40–60% quicker guide creation. Support teams, in particular, are finding they can offload 80% of their ticket volume with agents—but only if those agents have the content to be useful. 

“The agent without the content is useless,” Einav warns, noting that most enterprise documentation is either years out of date or entirely undocumented.

Community and industry early reception

Guidde already claims 4,500 enterprise customers and seeks to expand this number with its new round of funding. Support and operations leaders have been vocal about the platform’s ease of use. Christopher Cummings, VP of Client Experience at DocNetwork, highlighted its ability to provide “quick, personalized video responses to customer questions.”

 Meanwhile, Wren Cotrone, a Director of Customer Support, noted that “Once you set the branding the way you want, you can really zoom through this stuff.”

Ronen Nir, Managing Director at PSG, summarized the investment thesis: “Guidde is solving one of the biggest blockers to successful AI adoption: the knowledge infrastructure.”

Why this matters now

The paradigm shift from text-only LLMs to agentic video intelligence is the defining trend of 2026. Guidde’s Series B signals that the “ground truth” for enterprise agents will come from raw video observation, not static documentation.

By capturing how work gets done across 10s of millions of workflows, Guidde is building a dataset that few others possess. 

As Einav put it: “It starts with humans in the loop, and over time moves toward full autonomy.” For the modern enterprise, the map is no longer a static document—it’s a living, breathing video intelligence layer that guides both the workforce and the agents that support them.

AI chatbots have become embedded in the lives of American teenagers, according to a report published Tuesday by the Pew Research Center.

While the most common uses of AI among this demographic are to search for information (57%) and get help with schoolwork (54%), teens are also using AI to fill roles that would typically be occupied by friends or family. Sixteen percent of U.S. teens say they use AI for casual conversation, while 12% use AI chatbots for emotional support or advice.

Some teens may find solace in talking to chatbots, but mental health professionals are wary. General purpose tools like ChatGPT, Claude, and Grok are not designed for such uses, and in the most extreme cases, these chatbots can have life-threatening psychological effects.

“We are social creatures, and there’s certainly a challenge that these systems can be isolating,” Dr. Nick Haber, a Stanford professor researching the therapeutic potential of LLMs, told TechCrunch recently. “There are a lot of instances where people can engage with these tools and then can become not grounded to the outside world of facts, and not grounded in connection to the interpersonal, which can lead to pretty isolating — if not worse — effects.”

Pew’s survey also shows a discrepancy between teenagers’ self-reported AI usage and the extent to which their parents think they engage with this technology. About 51% of parents said that their teen uses chatbots, while 64% of teens reported using them.

The majority of parents are okay with their teens using AI to search for information (79%) or get help with schoolwork (58%), but far fewer parents approve of their teens using AI chatbots for casual conversation (28%) or to get emotional support or advice (18%). In fact, 58% of parents are not okay with their child using AI for such purposes.

AI safety is a contentious topic among leading tech companies, to say the least. But one popular chatbot maker, Character.AI, made the choice to disable the chatbot experience for users under the age of 18. This decision followed public outcry and lawsuits filed over two teenagers’ suicides, which took place after prolonged conversations with the company’s chatbots. OpenAI, meanwhile, made the decision to sunset its particularly sycophantic GPT-4o model, which sparked backlash from people who had come to rely on the model for emotional support.

Though a majority of teens use AI chatbots in some way, they have mixed feelings about the impact of this kind of technology on society. When asked how they think AI will impact society over the next 20 years, 31% of teens said the impact would be positive, while 26% said it would be negative.

It’s a blustery January day outside Lakehurst, New Jersey. The East Coast of North America is experiencing its worst weather in decades, and all civilian aircraft have been grounded the past four days, from Florida to Maine. For the past two days, that order has included military aircraft, including those certified “all weather” – with one notable exception. A few miles offshore, rocking and bucking in the gales, a U.S. Navy airship braves the storm. Sleet pelts the plexiglass windscreen and ice sloughs off the gasbag in great sheets as the storm rages on, and churning airscrews keep the airship on station.

If you know history you might be a bit confused: the rigid airship USS Arkon was lost off the coast of New Jersey, but in April, not January. Before jumping into the comments with your corrections, note the story I’ve begun is set not in 1933, but in 1957, a full generation later.

The airship caught in the storm is no experimental Zeppelin, but an N-class blimp, the workhorse of the cold-war fleet. Yes, there was a cold war fleet of airships; we’ll get to why further on. The most important distinction is that unlike the last flight of the Arkon, this story doesn’t end in tragedy, but in triumph. Tasked to demonstrate their readiness, five blimps from Lakehurst’s Airship Airborne Early-Warning Squadron 1 remained on station with no gaps in coverage for the ten days from January 15th to 24th. The blimps were able to swap places, watch-on-watch, and provide continuous coverage, in spite of weather conditions that included 60 knot winds and grounded literally every other aircraft in existence at that time.

Rigid? Count (Zeppelin) Me Out

Airships come in multiple types: rigid, non-rigid, and semi-rigid. Most people — my past self included — assume that the rigid type is more advanced. Unlike rigid airships, which are stabilized by an aluminum skeleton (or a wooden one, in the case of the Schütte-Lanz ships of the Great War), a blimp’s shape is maintained by gas pressure alone. Just a balloon with motors, if we’re being uncharitable. This limits the maximum speed, as the aerodynamic pressure of moving through the atmosphere increases with the square of the airspeed, and must always be lower than the internal pressure of the gas bag. You can’t even pressurize the gas bag much to compensate, because then the density of the lift gas gets too high to actually, well, lift.

Put a skeleton in there, and your airship can be much, much larger. It can go much faster. It can become a flying aircraft carrier, like the ill-fated USS Arkon, and its ill-fated sister ship, USS Macon. The U.S. Navy has only ever fielded five rigid airships; only one survived long enough to be decommissioned. It is with no disrespect to the brave men and women who served– and lost their lives– aboard those silver giants that we dismiss them from our narrative here. They were a worthy experiment, but a failed one. By contrast, the U.S. Navy fielded 166 blimps in the Second World War, and only a handful were lost, mostly during ground handling, and one to enemy action.

So, how was an N-class blimp, also known as a ZPG-2, in the designation system of the day, or SZ-1A after 1962, able to ride out a storm much worse than the one that sank its rigid-framed predecessors? It’s probably precisely because it lacked that rigid frame. The non-rigid envelope of the blimps could bend, buckle, twist, and alter their shape in response to strains that would break the keel of a Zeppelin. Non-rigid airships can quite literally flex on their rigid cousins when it comes to airworthiness.

The flexing skin of a blimp turns the entire gas-bag into one giant de-icing boot to boot, keeping yet another weather hazard at bay. Icing is a great danger to aircraft: when conditions are just wrong, like during the January storm described above, it’s easy for the weight of ice to build up and bring down any aircraft without an effective de-icing system. De-icing boots are one such system: rubber membranes, typically on the leading edge of the wing and tail surfaces of an airplane, that are inflated to flake off ice. On airplanes, they’re addons, but it’s a built-in bonus to flying a blimp.

Of course another key advantage of non-rigid airships is that they’re just plain cheaper. Being smaller, they require less crew, less ground crew, and smaller hangers, but a small rigid would have the same advantage. More importantly, especially during wartime, is that a Zeppelin requires everything you’d use to build the equivalent blimp, plus all the Duraluminum (or other material) going into its rigid frame. Logistically speaking, blimps were a no-brainer if the US wanted to field a lot of airships, and at one point they certainly did.

But Why?

Unlike a certain (in)famous penguin, the US Navy knew exactly what it was doing when it ordered the N-class airships after World War Two. As stated, they had over a hundred blimps in service during that conflict, and racked up more Lighter-Than-Air (LTA) flight time than any other organization has before or since: 550,000 hours split over 55,900 sorties in both the Atlantic and Pacific theaters. While the institutional knowledge is long gone, it’s safe to say that in those days nobody knew airships like the U.S. Navy knew airships.

The vast majority of the wartime fleet — some 135 examples — were of the K-class. These ships were designed with a specific mission in mind: antisubmarine warfare. Blimps vs subs wasn’t a new idea; the Americans had worked with the Royal Navy’s u-boat hunting blimps in the First World War. Though the Royal Navy gave up on the idea after the conflict, interest remained on the other side of the Atlantic, and history shows the Yanks were right to persist with it. Of roughly 89,000 ships in blimp-escorted convoys, only one, the tanker Persephone, was sunk, ironically off the coast of New Jersey, not terribly far from the Lakehurst home of LTA.

The sub-hunting blimps were perhaps making it up as they went along. On paper, though, the airship is ideal for the role: without needing to burn fuel to stay airborne, it can have absurdly long loiter times. Its low speed is of no issue when shadowing convoys that have to move at the speed of the slowest merchant vessel– even the HX series “fast convoys” didn’t exceed 13 knots (24 km/h). Blimps of the K-class could cruise at 50 kn (92 km/h), and dash at up to 68 kn (125 km/h), which proved more than sufficient to keep up.

When the class was designed in 1937, its ability to cruise low and slow was ideal for hunting submarines with the Mk.I eyeball, but by the time the K-class was fielded in numbers in 1942, they were also equipped with first-generation radar, magnetic detection coils, and even primitive sonoboys after 1943. The class proved flexible and continued to be upgraded with the latest equipment until the last “K-ship” was retired from active duty in 1959.

At 251 ft 8 in (76.73 m) long, with a gas-bag diameter of 57 ft 10 in (17.63 m), the K-ships could lift a crew of 9 in relative comfort, with fuel to feed their twin Pratt & Whitney R-1340 radials for 38 hours of normal operation. Idling the engines and making use of air currents could extend that number by quite a lot compared to cruising steadily, of course. As stated above, in wartime the K-ships carried magnetic detectors, sonobouys and radars for U-boat detection, along with four depth bombs and a .50 cal machine gun for weapons.

If four bombs doesn’t sound like much, well, that’s probably why no U-boats were recorded killed by Navy airships. On the other hand, the main mission of the blimps was to protect convoys, not to sink subs. “Damaged and driven off” was good enough, especially when the blimp could track the wounded u-boat from above and direct other assets like destroyers to make the kill, as often happened. There was a larger M-class designed during the war that was half again the size of the K-ships and could thus carry eight depth charges, but only four were built before the conflict ended.

Post-war, one K-ship by the name Puritan was sold back to Goodyear and equipped with 1,820 incandescent light bulbs to serve as a floating ad ticker, which perhaps shows the versatility of the design. Alas, ad revenues did not cover the cost of keeping the 425,000 ft³ (12,035 m³) envelope filled with precious helium. Civilian blimps since have been of more modest size.

The LTAs that Aren’t

Speaking of precious helium, in order to conserve that lift gas, the Navy actually operated their blimps as Lighter-Than-Air craft as little as they possibly could, both during and after the war. An annoying thing about airships is that they get lighter the longer they fly as they run down their gas tanks. It is possible to run an engine on a hydrocarbon gas with a density similar to air, like the “blau gas” used by the Graf Zeppelin in the 1920s, but this has one major drawback: it’s a major logistical headache to require a special fuel for a relatively small number of units. Though there was one prototype with a blau gas style fuel in the 30s, the US Navy put logistics first. For the war and several years afterwards, everything that the Navy flew would burn AvGas, at least until the jet age made things annoyingly complicated for quartermasters.

Without special fuel, the issue of excess lift can be mitigated by condensing water from the exhaust, but that doesn’t quite balance out, so the problem still remains on long flights. Eventually one must either vent helium to reduce lift, which is wasteful, or take on ballast to make up for lost mass, which can disrupt operations. The alternative the US Navy preferred was to fly “heavy”.

Yeah, it turns out hybrid airships– craft that combine lift gas with aerodynamic lift–aren’t a new idea. You might not think of the teardrop-shaped gas bag of a classic blimp as an airfoil, but with a little airspeed just a modest nose-up attitude– what a pilot would call ‘angle of attack’–the blimp can get considerable dynamic lift. By accepting the tradeoff of requiring a takeoff run, the blimps could get into the air with enough dynamic lift to account for the expected fuel burn, and come back to base with only so much lift capacity that could be cancelled out by trimming the ship downwards.

The Cold War Era

After the war, most of the K-ships were crated-up and decommissioned, and their air and ground crews were amongst the first to be demobilized. “Most” does not mean “all”, and once the thrill of peace turned into the uneasy truce of the Cold War, Uncle Sam was glad to have those airships. The Soviets had submarines, too, after all.

Rather than continue with building more of the M-class, the decision was made to update the existing stocks and produce improved K-class ships for the immediate post-war period. The wartime ships that were not decommissioned were updated with better electronics and a 20% larger gas bag, getting the designation ZPK2 and then a further upgrade to ZPK3 standard. Fifteen new K ships were built by Goodyear after the war and delivered starting in 1953 under the designation ZPK-4. The last revision of that design, ZPK-5, was built with an inverted “Y” tail instead of the standard cruciform to allow for greater nose-up attitude during the ‘heavy’ takeoffs mentioned above. Twelve ZPK-5s were built by Goodyear and delivered from 1955.

While the K-class was being modernized with better sensors and weapons, the US Navy’s LTA program recognized that it could not simply coast on legacy wartime engineering.They therefore commissioned Goodyear for a clean-sheet design that would be another 50% larger than even the four M-ships, which were kept in service until 1956. These new airships would become the N-class whose all-weather adventures this article opened with.

While the ZPG-2W whose triumph we described above were built to serve the airborne early warning role, most– twelve out of seventeen–of the “Nan ships”, as the class was called, were initially designed as bigger, badder sub-killers in case war broke out with the Soviets.

They had better down-looking radars– the AN-20, the best available at the time–much improved sonobouys, more sensitive magnetic anomaly sensors, and homing torpedoes. In war games against US and allied diesel-electric subs, like the GUPPY class, they proved very effective indeed, as did the improved K-ships. Against the new, nuclear-powered USS Nautilus, they were much less successful, but so were fixed-wing and helicopter assets. Doctrine that relied on spotting subs while recharging at snorkel or on surface was ill-suited to deal with a ship that could run submerged for months.

Improving on the control arrangement of the ZPK-5s, the Nan ships were built with an X-shaped tail to allow for even greater pitch angles during takeoff without tailstrikes. The ruddervators on the X-tail could also be controlled by one pilot, compared to earlier blimps which needed separate operators for elevator and rudder. The largest difference in design was perhaps the buried engines: unlike previous Navy blimps, which used radial engines hung from the gondola, the ZPG-2 Nan ships kept their two 800 HP Wright Cyclones indoors. This was supposed to allow for maintenance during flight, and it allowed the engines to be coupled together via a clutch, allowing single-engine cruising. As the air-early-warning blimps proved in 1957, these were all-weather craft.

The Anti-Submarine Warfare (ASW) squadrons gave a similar demonstration in 1960 with “Operation Whole Gale” during which the Nan ships provided 24/7 coverage for two full months, again in the teeth of winter’s worst weather. In spite of their best efforts to make use of wind and storms, no submarine got past the blimps during the operation.

The post-war record of the US Navy’s blimps is full of such impressive moments. The service was very much looking to prove itself, and so jumped at opportunities to demonstrate the blimps’ capabilities. Arctic expeditions? A Nan-ship proved its worth on 24-hour patrols between Churchill, Manitoba and Resolute, Baffin Island– the last airship to cross the Arctic Circle. Another stunt in 1957 set a record for unrefueled flight: a circumnavigation of the Atlantic basin from Massachusetts to Portugual, North Africa, and finally ending in Florida that took 264 hours and spanned 9,448 nautical miles (17,500 km). Guinness will tell you that Graf Zepplin’s 71-hour 6,384.5 km trip from Fedrickshaven to Lakehurst holds the record for airship flight, but that’s seriously out-of-date. For a rigid, sure, that’s the record, but for any LTA? Blimps win. Blimps actually win all the airship records save for speed and size, and none of those records stand from the “golden age” of the 1930s.

Takeways

That’s maybe the lesson here. Blimps win. I consider myself something of an aviation geek, and have multiple books on airships. All of them tell the same story: blimps were a sideshow, Zeppelins were the pinnacle of airship engineering, and it all ended with the Hindenburg. That’s the story everyone knows, just like everyone knows that airships are useless in any kind of bad weather.

What everyone knows is wrong. The problem with the story we all know is that it ends 24 years early, and leaves out more flights than it includes. Add in those 24 extra years of innovation, and the blimps come off looking a lot better in comparison.

The last flight of a US Navy dirigible with a US Navy crew was in August 1961. The ZPG-2 Nan ships were followed by a larger ZPG-3: bigger again, with a larger, more capable AN-70 radar hiding in the gasbag, the ZPG-3 was the largest blimp ever fielded. Its capability didn’t matter– there was no money for blimps. Imagine a line of Admirals standing before the US Congress, hats in hand, and one asks for money for nuclear-powered submarines to smite the enemies of Uncle Sam with Intercontinental Ballistic Missiles wielding atomic fire, and the next man in line wants money for blimps. Airships seemed positively old-fashioned in comparison, and money was tight. The blimps were cut.

Yes, they provided an all-weather ASW and AEW capability nothing else could match… but other assets, ships and airplanes and helicopters, could do 90% of the job without requiring the expensive, dedicated infrastructure the blimps did. Airships were cut from the U.S. Navy the same year as seaplanes and the Regulus cruise missile program. You might say they’re the only things ever destroyed by the Polaris missile subs, but that’s arguably a good thing.

All the hot venture capital money is being sucked up by the AI bubble right now, and even if it wasn’t, the trendy thing in aviation is electric vertical takeoff and landing. That doesn’t mean there isn’t an airship renaissance just around the corner– there is always an airship renaissance just around the corner. That it never results in anything but prototypes is irrelevant. LTA is just too enticing a technology to ever give up. If we ever are to get that renaissance to bear fruit, though, we’re going to have to have better stories.

If you’re focused on the Hindenburg going down in flames, or the Akron and Macon breaking up over water, airships seem like a bad bet. If you remember the Nan ships bouncing and wiggling their way through January snowstorms when everyone else was grounded, then LTA starts to sound more reasonable.

It’s the classic awkward icebreaker: If you could invite anyone, dead or alive, to a dinner party, who would it be? Aristotle? Ailias is a company based in Surrey, UK, which promises to make that hypothetical a reality. It can reanimate historical and current legends with 3D hologram avatars that are fully conversational, knowledgeable, and can be delivered to you in a box.

The technology isn’t bespoke. Many companies provide life-size hologram displays for events and parties, everything from floating 3D displays of Santa’s sleigh or 3D Holo-Trucks. The physicist Dennis Gabor even won a Nobel Prize in 1971 for his work that led to holography, even though a life-size Elon Musk isn’t probably the result that he (or anyone) had in mind.

What sets Ailias apart is the company’s playful focus on history and education, which the company describes as “ultra character creation.” The company focuses on animating dead notable personalities into real-feeling conversational holograms, designed for interaction rather than spectacle. Ailias’ holograms can juggle, do squats, or even breakdance, making your party, exhibition or just about any event an extra special occasion.

Man in the Box

Ailias offers pricing on request, with costs varying depending on whether clients opt for rental, purchase, or whether you’re seeking bespoke characters and activation. When I visited the offices, director Adrian Broadway noted that a minimum week’s rental would run into the thousands of pounds, which includes software subscription costs, delivery, and installation.

Ailias’ current roster has over 70 characters that could be staged in their bespoke boxes, including Henry VIII, Beethoven, Julius Caesar, and a suspiciously sexy Cleopatra. That these are mostly historical figures is no coincidence—Broadway describes these boxes as great for educational settings or museum exhibitions, but admits it also has to do with copyright restrictions on characters as well.

In the United Kingdom, the use of someone’s identity for commercial purposes is treated as a trademark. (In the United States, the right to publicity is protected in some form in most states.) That is to say, if Ailias used a well-known or living celebrity, that would likely land the company in court. But a long-dead historical figure like Henry VIII is unlikely to cause trouble.

In this instance, Ailias had cleared the copyright concerns for the 7-foot-tall AI Albert Einstein, so after hitting the Start Chat button, I talked to Einstein about a wide range of topics, everything from science, music, to his thoughts on Elon Musk. He had a pleasant, soft German accent, and I was impressed at the response speed. Ailias notes that it takes under two seconds for each avatar to respond, which feels about right.

Photograph: Dulcie Godfrey

For an educational hologram, I often found myself answering more questions than I was asking. There were times Einstein felt like a large, animated ChatGPT conversation but with a German accent. This is to be expected, as Ailias relies on open source AI and third-party generative video to create the conversations. But there’s no sense of verisimilitude anyway, since Einstein wasn’t really 7 feet tall. I took the opportunity to ask, like an 11-year-old boy would, “Who would win in a fight, you or Isaac Newton?”

It held up as any AI language model would, deflecting back to its area of expertise by settling on a sensible, “It would be more of a fight of ideas.” In the aim of being at least semi-professional, that’s as far as I went. But I’d imagine the language model would do fine with most things a preteen could throw at it.