This Week in Tech (TWiT) – Episode 1025: "Weak Perfection – Signalgate, Miyazaki's Nightmare"
Date: March 31, 2025
Host: Leo Laporte
Guests: Alex Kantrowitz (Big Technology), Ian Thompson (The Register), Jacob Ward (The Rip Current)
Overview
Episode 1025 dives deeply into an eventful week in tech news, tackling the explosive AI-driven "Studio Ghibli-mania," OpenAI’s PR tactics, and the ethical minefields of generative art. The panel debates the White House Signal breach ("Signalgate"), examines the fallout from 23andMe’s bankruptcy, and speculates about the future of creative professions and privacy in the AI era. With a mix of humor and urgency, the show also reflects on technological acceleration, the blurry lines between safety and progress, and how the tech world is converging with politics and society at large.
Key Topics & Discussion Points
1. Studio Ghibli, AI Art, and the "Miyazaki’s Nightmare" Meme ([06:26]–[31:07])
- Emergence of AI-Generated Ghibli-Style Images:
OpenAI added image creation to ChatGPT-4o; a flood of Studio Ghibli–style content (especially memes) quickly captured the internet’s attention. - Viral Demand and Limits:
Servers were overwhelmed by the demand, with CEO Sam Altman pleading for users to "stop creating images—you're burning down the servers" ([12:17]). - Ethics of Artistic Appropriation & Impact on Creators:
- Jacob Ward: "You're grabbing Miyazaki, the most beloved...if you're a young person who likes to be alone in your room and live in a calm world...to have this as the way OpenAI shows off this latest twist in this technology...it's just the most tone-deaf version again" ([09:27]).
- Alex Kantrowitz: "I think they lean into the controversy as a marketing tool, and it’s absolutely working for them" ([11:56]).
- Ian Thompson: "I think AI is just going to turbocharge [the lack of new original films] completely. I’m sure studios looked at it and thought, 'Yes, we can get rid of those pesky actors...'" ([16:53]).
- Fandom vs. Exploitation:
Are AI-generated fan works a natural continuation of historical fandom ("derivative works"), or do they hasten the end for handcrafted original IP? - Will AI Replace Human Creativity?
- Leo Laporte: “While I’m not a huge fan of CGI, it’s in every movie… it changed filmmaking without destroying it” ([18:37]).
- Jacob Ward: “Just give them two financial quarters and they'll figure that stuff out. The quality is going to go up and up…” ([21:17])
- Copyright & Legal Landscape:
Generally, mimicking a visual style is not copyright infringement, per IP attorneys ([20:46]).
Notable Quote
"I'm seeing these animators on social media literally weeping into the camera, saying, Like, 'This is the end of it. This is it.'" —Jacob Ward ([13:38])
2. The Signal Breach – "Signalgate" ([53:56]–[63:56])
- Leaked Chat with Top Officials:
White House’s use of Signal for sensitive, arguably illegal, discussion of attack plans. A journalist (Jeffrey Goldberg of The Atlantic) was inadvertently included, leading to a blockbuster scoop. - DARVO Tactics in the Aftermath: (Deny, Attack, Reverse Victim and Offender)
- Jacob Ward: Explains how officials deflected and denied responsibility: "It is the language of politics now, and the response to this is exactly that" ([55:05]).
- Security Flaws and Practical Risk:
Signal messages, while encrypted in transit, are decrypted on device—compromised phones pose a risk even with secure apps: “Signal is totally secure, but individual handsets aren’t. That’s the key.” —Ian Thompson ([63:34]) - Implications for Policy and Practice:
- Using Signal is (likely) illegal for such communications; courts rule records must be preserved
- Calls into question governments' digital hygiene, the challenge of enforcing secure communications, and a generational shift in policy adherence
Notable Humorous Moment
"My suspicion... Some idiot just picked the wrong J.G." —Ian Thompson, speculating on a likely mundane source for the screw-up ([58:16])
3. AI Job Displacement and the Evolution of Human Work ([39:09], [125:13])
- AI's Increasing Capabilities & the Job Market
- Alex Kantrowitz shares personal doubts: “Last year...ChatGPT didn't take my job. This year, I'm starting to see things that are making me nervous” ([39:14]).
- Law, Coding, Creative Jobs’ Uncertain Future:
Lawyers, coders, and even journalists mostly agree: AI isn’t there yet, but it has encroached further and faster than predicted. - Unusual Jobs of the Future:
Creative, hospitality, and “human-touch” professions may be safer—at least for a while ([129:23]).
- Human + AI Collaboration:
Human creative "10%" still matters ("the magic"), but persistent dissatisfaction with being an 'AI-overseer' role ([25:52]).- Jacob Ward: “The top 10% who are using it the most reported deep dissatisfaction with the job. They hated being in the role of catching the AI’s work...” ([25:43])
- Societal/Economic Spiral:
- Can billion-dollar companies with one employee sustain a consumer marketplace if no one else has money?
- “If you fire them all, they won’t have any money to buy it.” —Leo Laporte ([44:15])
4. Government Tech Modernization and Social Security Software ([93:00]–[105:28])
- Elon Musk’s DOGE & Government Overhaul:
- Plans to completely replace Social Security’s 60M-line COBOL code base in six months with AI’s help ([94:28])
- Skepticism on Speed and Safety:
- “If you do break Social Security, that's a massive political liability...” —Alex Kantrowitz ([105:28])
- Leo and panel strongly argue for caution: “You would do it in parallel. … You WOULD take your time to do it right, because an error … could really be a life or death situation.” ([98:37])
- COBOL Defended:
Long-lived infrastructure isn't inherently bad if it just works: “If it’s working, you don’t need to rewrite it.” —Leo Laporte ([97:17])
- Broader Problem:
- U.S. government software is “awful, miserable, makes employees less efficient,” largely due to procurement policies. Fixes are complex—requiring more than just moving fast and breaking things.
5. Data Privacy, 23andMe Bankruptcy, & DNA for Sale ([113:55]–[123:54])
- 23andMe’s Collapse & Data Retention:
- Company files for bankruptcy; CEO Anne Wojcicki steps down (with plans to potentially buy the assets back)
- California Attorney General recommends deleting your data
- Past hacks: 7 million records breached, lawsuit settlements
- The Dangers of Genetic Data:
- Jacob Ward: Explores how data can drive “genoeconomics”—predictive modeling for things like educational achievement, which could fuel discrimination and eugenics ([115:35]–[121:12])
- Once compromised, DNA can't be changed: "You can’t change your DNA the way you change your password." —Jacob Ward ([121:07])
- Data increasingly used for things like insurance profiling, law enforcement, or more sinister purposes
- Benefits and Ethical Complexity:
- Success stories of family reunification and cold cases solved via genealogy; but—"minimize your data footprint," urges Ian Thompson ([121:48])
Quotes & Memorable Moments
-
On AI’s Marketing Style:
“I think they lean into the controversy as a marketing tool and it’s absolutely working for them.” —Alex Kantrowitz ([11:56]) -
On Generative Art's Impact:
“We don’t have anything new to sample, we’re just rehashing the same stuff.” —Jacob Ward ([16:53]) -
On Industry Disruption:
"Disney can’t make an animated feature for less than $200 million. ... Is that a threat to studios?... Maybe the best creative output will rise." —Alex Kantrowitz ([17:27]) -
On Security Breach Etiquette:
“If we did this, we'd be in Leavenworth serving time. Why is the Secretary of State for Defense allowed to get away with it?” —Ian Thompson ([62:47]) -
On the Impossibility of Catching Up:
"Give them two financial quarters and they'll figure that stuff out." —Jacob Ward ([21:17]) -
On the Loop, Fast/Slow Thinking, and AI:
"Our brains are shortcut machines... AI is the perfect kryptonite for that." —Jacob Ward ([137:17]) -
On Privacy in a Face Recognition World:
"This is the world we're entering now. There's no anonymity because of face recognition." —Leo Laporte ([145:07]) -
On Historical Tech Cycles:
“This is what happens with technology—disintermediation.” —Leo Laporte, on the fate of history websites amid AI aggregation ([133:24])
Segment Timestamps
- [06:26] – Studio Ghibli, ChatGPT Image Generation, and Meme Culture
- [13:28] – Human vs. AI Creativity: Industry Turmoil
- [39:09] – The Looming Impact of AI on Jobs; Journalism & Customer Service
- [53:56] – The White House Signal Breach: Security, Ethics, and Embarrassment
- [63:56] – Why Signal Isn’t Enough for National Security
- [93:00] – Musk, DOGE, and the Mad Rush to Rewrite Social Security in Months
- [113:55] – 23andMe Bankruptcy: Should You Delete Your Data?
- [125:13] – What Should Young People Study? AI, Art, and Safe Professions
- [129:45] – The Future of Human Work and Content Aggregation
- [145:07] – Face Recognition, Privacy, and the MSG Scandal
- [150:53] – Gemini Pro Goes Free, Economics of AI Access
Episode Tone & Style
Ranging from wry and humorous (“Why am I blue, and why is Ian a dog?” [34:51]) to alarmed and urgent about the pace of technology, the discussion remains conversational but pulls no punches when confronting the tech industry's pitfalls, hype cycles, and ethical lapses. Frequent real-world examples, references to classic legal and economic principles, and personal anecdotes keep the pace lively.
SUMMARY
“Studio Ghibli-mania” via AI has stirred a cultural hornet's nest, thrusting questions of artistic ownership, fandom, and technological overreach into the spotlight. OpenAI’s controversial marketing and strategic “leaks” have both stressed their infrastructure and amplified the debate about the ethics of AI-generated art. Meanwhile, the “Signalgate” White House breach is a cautionary tale about “shadow IT”—security is often about people, not just tools.
Rapid advances in AI are raising existential questions for creative and analytical professions and driving a wedge between the market’s relentless efficiency and the values we risk discarding. The episode consistently pushes listeners to ask: Who benefits, who is at risk, and what kind of future are we racing toward? As AI bulldozes old boundaries, the panel collectively—if sometimes uneasily—embraces the need for constant debate, critical thinking, and a willingness (as Leo says) “to keep talking about it.”