Episode Summary
Executive Summary: Hard Fork’s New Year episode revisits 2024 predictions and 2024 resolutions, then lays out 2025 tech forecasts, including AI culture-war backlash, meme-coin speculation, Waymo’s mainstreaming, a possible Apple-Snap deal, X’s integration into xAI, and OpenAI declaring AGI. The second half answers listener questions on VPNs, podcast authenticity, AI image generation limits, training data, voice-clone security risks, and streaming-account ethics.
Main Topics: 2024 prediction postmortem (Priority: 5/5): Kevin and Casey grade last year’s forecasts, finding mixed success on Threads vs. X, a miss on “lawless” LLM adoption, partial accuracy on Google/Gemini catching up to OpenAI, a miss on white-collar unionization over AI, and a strong call on X’s Hunter Biden-style censorship scandal. 2025 tech and AI predictions (Priority: 5/5): They make new forecasts: Casey predicts an AI culture war will erupt, Kevin predicts a crypto meme coin will briefly hit $100B market cap, Casey predicts Waymo will go mainstream, Kevin predicts Apple will acquire Snap, Casey predicts X will merge into xAI, and Kevin predicts OpenAI will declare AGI. New Year’s resolutions and attention management (Priority: 4/5): They review 2024 resolutions about reducing doomscrolling and improving YouTube habits, then set 2025 intentions: Casey wants to use AI to build a meditation practice, while Kevin wants to re-enter social media more actively and constructively instead of passively criticizing it. Listener questions on VPNs and platform behavior (Priority: 4/5): Whitney Jones joins to answer reader mail on why VPNs are used, how they work, and where they’re legitimate or illicit, emphasizing privacy, geo-unblocking, secure corporate access, and political use in authoritarian contexts. AI limits, training data, and safety (Priority: 4/5): The show explains why image generators fail on odd prompts like “a loon in a bathtub,” why video data is increasingly valuable for AI training, and why voice-verification systems are vulnerable to AI voice cloning. Authenticity and ethics in digital media use (Priority: 3/5): They address concerns about whether the podcast is AI-generated, note the rise of NotebookLM-like audio styles, and debate whether it’s acceptable to use a former partner’s streaming login—ultimately treating free viewing as mostly harmless but cautioning against purchases and shared watch-state confusion.
Key Arguments: Threads likely grew substantially in 2024, but platform metrics are too fuzzy to resolve a clean win over X; X remains culturally more influential despite its decline. AI chatbot usage did not clearly migrate toward “lawless” models, though open-source and lightly restricted models like Llama and Grok have gained traction. Google improved model quality and Gemini may top some benchmarks, but ChatGPT still dominates cultural awareness and apparent usage. White-collar AI-related union mobilization did not materialize in 2024, suggesting the labor response may come later. The Vision Pro did not become a mass hit, but mixed reality interest shifted toward Meta Ray-Ban-style glasses and broader smart-glasses competition. A real AI culture war is likely in 2025, with congressional scrutiny, partisan backlash over chatbot behavior, and debates over intimate chatbot relationships. Crypto meme coins and gambling-style speculation are poised for a comeback under a crypto-friendly political environment. Waymo’s expansion into more cities could make autonomous driving a mainstream cultural object, especially once it reaches New York. Apple may have strategic reasons to buy Snap, but privacy/safety concerns around teen communication could deter the deal. If OpenAI wants to break dependence on Microsoft and reclaim control, it may declare AGI once a new model launches and trigger contractual consequences. Voice-based authentication is insecure and should be replaced, since cloned voices can defeat it with only short audio samples.
Data Points: Threads monthly users: 275 million - Casey cites Threads’ reported user base while arguing it had a big 2024. Threads app-store ranking: #2 in the U.S. - Threads is described as near the top of the U.S. App Store for about a month. Meta AI assistant monthly active users: nearly 600 million - Mentioned as a proxy for Llama-based usage, though both speakers express skepticism about Meta’s numbers. ChatGPT weekly users: 300 million weekly users - Used to argue that ChatGPT remains the dominant consumer chatbot. Apple Vision Pro sales estimate: about 420,000 units - Referenced as a rough 2024 sales estimate for the headset. Vision Pro revenue estimate: about $1.5 billion - Presented as enough first-year revenue to justify another product iteration at most companies. Hawk meme coin peak market cap: almost $500 million - Used as evidence that crypto meme coins can rapidly inflate and collapse. Hawk meme coin crash: more than 95% of value lost within hours - Illustrates the speculative, pump-and-dump nature of meme coins. XAI valuation: $50 billion - Used to support the idea that Elon Musk may increasingly prioritize xAI over X. Twitter acquisition valuation: $44 billion - Compared against xAI’s valuation to highlight how valuable AI is perceived to be. VPN market size: well in excess of $40 billion - Cited to show VPNs are a large, established market. UK/authoritarian VPN trend: VPN downloads in Russia went through the roof - Used to show political suppression drives VPN adoption. AI image prompt example: common loon 26–28 inches; standard bathtub about 60 inches - Claude generated a rough proportionality check for the “loon in a bathtub” prompt.
Pivotal Quotes: "I want to get somebody who's in severe legal jeopardy to come on the hard fork and just air it out." — Kevin Roose: Opening banter about wanting more controversial guests under active investigation. "I think over the next year, as chatbots sort of enter more and more facets of Americans' lives, we're going to start to see the rumblings of a backlash here." — Casey Newton: Casey explains why he expects an AI culture war in 2025. "I would like to get medium good at meditation using AI." — Casey Newton: Casey’s 2025 personal resolution, blending AI with self-improvement.
Implications: Listeners should expect 2025 to bring more AI controversy, more speculative crypto behavior, and more visible autonomous vehicles, while also facing sharper questions about privacy, authenticity, and safe use of consumer AI tools.
About Hard Fork
“Hard Fork” is a show about the future that’s already here. Each week, journalists Kevin Roose and Casey Newton explore and make sense of the latest in the rapidly changing world of tech. Unlock full access to New York Times podcasts and explore everything from politics to pop culture. Subscribe today at nytimes.com/podcasts or on Apple Podcasts and Spotify. Also, for more podcasts and narrated articles, download The New York Times app at nytimes.com/app.