(Preview) Sora 2 and an AI Video Boom, Meta’s Vibes One Week Later, Questions on TikTok, Solar Power, and the iPhone Air
Summary
Sora 2 plausibly clears the more defensible “GPT-3.5 moment” bar: not a new ChatGPT-scale event, but the point when generated video becomes a coherent, consumable finished thought. Ben Thompson initially “poo-pooed” OpenAI’s comparison, then distinguished GPT-3.5 from ChatGPT itself: short clips now hold together well enough to astonish mainstream users, even if experts have already seen related capabilities from Google’s Veo models. “It’s sufficiently finished to be something meaningful like GPT-3.5 was.”
Andrew Sharp sees ChatGPT-like virality but not ChatGPT-like utility. Sora 2’s photorealistic output, intuitive interface, and unsolicited excitement from people outside the early-adopter tech crowd feel unusually consequential, yet Sharp wants to leave after roughly 90 seconds and cannot identify a daily-use case beyond entertainment. His pushback: ChatGPT was immediately useful, while Sora currently looks “useless”—a condition he explicitly says might change.
OpenAI’s repeatable advantage may be productization, not merely model quality. ChatGPT, the Studio Ghibli image-generation moment, and now Sora represent three viral launches; Thompson’s investor-relevant conclusion is that “three times is worth keeping track of.” Sora reached No. 3 in the App Store while Meta AI was roughly No. 125; Gemini was No. 1, albeit invite-gated.
AI video could expand the creator class enough to produce an order-of-magnitude content shock. Thompson’s 1%-create, 9%-distribute, 90%-consume model may break when generating video stops being a “big pain in the ass”: moving creators from 1% to 2% already doubles supply, while a tripling or even 10× increase could surface ideas previously blocked by production constraints. The structural change is “the breaking apart of the idea and the substantiation of the idea.”
Meta’s attention franchise faces a potentially disruptive format it cannot simply clone inside Instagram. Thompson says Meta’s power and advertising business are predicated on attention. TikTok took attention, but Meta could answer with Reels; Sora presents two harder problems if momentum persists—Meta’s AI may not be good enough, and users may not want AI video shoved into Instagram. His call is conditional but blunt: “Meta might be in big trouble here.”
The hostile reaction to Meta Vibes exposed narrative—not visual novelty—as the missing ingredient for most users. Listeners criticized Vibes as neither funny nor sexy and as lacking narrative; Sharp added that it also offered no apparent information or instruction. Thompson’s own screensaver comparison reflects his atypical taste, and he conceded that treating ambient, narrative-free imagery as broadly compelling was “a massive blind spot.”
Sharp treats today’s AI-video apps as half-baked prototypes whose destination is still unclear. He compares them with computers in 1993 or 1994: an interesting glimpse of where things might be in four or five years, even though he largely agrees with the Vibes backlash. Thompson revisits his optimism about AI-generated worlds, acknowledging that fantastical environments may attract a committed niche without becoming broadly compelling.
A recurring outside critique is that frontier resources are being directed toward attention decay. Andrew read Ashkay Kothari’s question about why people keep dedicating “our brightest minds, billions of dollars, and the most powerful GPUs on Earth” to another app optimized for attention decay, noting that versions of this criticism had recurred about both Vibes and Sora.
Deep dive
1. Sora 2 and the finished-thought threshold
Thompson initially “poo-pooed” OpenAI’s GPT-3.5 comparison as an audacious attempt to invoke a seminal technological moment, then corrected himself: OpenAI claimed a GPT-3.5 moment, not a ChatGPT moment. That distinction changed his judgment from skepticism to qualified agreement.
His historical framing: GPT-1 and GPT-2 showed emergent capability, while GPT-3 could make sense before trailing into “nothingness.” GPT-3.5 remained flawed and hallucination-prone, but it could return a cogent response with a beginning, middle, and end—something ordinary people could consume.
Sora 2 now does the analogous job for video. Its short clips form “a finished thought,” unlike the original Sora, which Thompson placed nearer GPT-1 or GPT-2 despite its striking demonstrations and the attention attached to ChatGPT’s maker.
Sharp’s representative feed captured the leap through absurd narrative: Ronald McDonald in a hamburger car during a high-speed chase on Los Angeles’s 110. The content was silly, but its visual coherence and accessibility made the technical advance immediately legible.
2. Mainstream surprise matters more than expert surprise
Thompson’s self-diagnosed blind spot is prolonged exposure to the frontier. Because Veo, Veo 2, and Veo 3 had already shown him high-quality generated video, Sora 2 did not feel shocking; people downloading an easy consumer app instead reacted with “Holy cow.”
DeepSeek supplied his analogy. Paying OpenAI users had already experienced thinking models, but many free users first encountered visible reasoning through DeepSeek’s “cutesy” presentation and saw a remarkably better product—not merely a novel Chinese app.
Sharp, the podcast’s “resident normie,” found Sora unusually significant: the interface was intuitive, the output largely photorealistic, and several people outside the early-adopter tech crowd independently texted him about it. Nothing else in AI over the prior two years had felt as close to ChatGPT-style consumer momentum.
3. OpenAI’s repeated product launches could expand content supply
Sharp’s strongest positive read was product execution: Veo and Meta Vibes existed, but Sora made creation feel obvious. Thompson agreed and counted three viral OpenAI moments—ChatGPT, the Studio Ghibli image-generation moment, and Sora—arguing, “This isn’t a thing that happened once.”
Their disagreement centered on utility. Sharp uses ChatGPT daily for household projects and toddler questions but bails out of Sora after about 90 seconds; a friend stayed until 2:00 or 3:00 AM, yet Sharp still saw no clear path from entertainment novelty to everyday tool.
Thompson’s pushback is worth keeping: entertainment-only products are routinely dismissed as toys, even though “the most valuable companies” capture value by entertaining people. Sora’s lack of productivity might therefore strengthen, rather than weaken, its commercial relevance.
More consequentially, AI could alter the 1%-create, 9%-distribute, 90%-consume ratio. If easier production moves creators from 1% to 2%, supply doubles; a tripling or possible 10× expansion would separate ideas from the difficult work of substantiating them and unlock creators previously constrained by video production.
4. Sora creates an awkward attention problem for Meta
Thompson still defended one part of his Vibes thesis: an all-AI feed lets viewers relax because every clip is knowingly fake, avoiding the cognitive tax of deciding whether each photorealistic video is authentic. He thinks a distinct AI feed may therefore be easier to consume than synthetic media mixed into an existing feed.
Sharp prefers Sora’s photorealistic, real-world output; Thompson finds that less interesting and prefers Vibes’ futuristic environments. That disagreement reinforces Thompson’s uncertainty about whether AI-generated worlds will appeal broadly or mainly to a niche.
OpenAI also avoided disrupting ChatGPT’s low-friction conversational interface by launching Sora separately. Meta’s opposite choice made strategic sense as an attempt to drive Meta AI installation, but Sora nevertheless reached No. 3 in the App Store while Meta AI was roughly No. 125. Gemini was still No. 1, albeit gated by invites.
Meta historically answered TikTok by putting Reels inside Instagram. If Sora keeps taking attention, Thompson sees two structural obstacles: Meta’s AI “just isn’t good enough” to match it, and users may not want AI video shoved into Instagram. “Attention is a precious thing.”
5. Vibes struggled because spectacle without narrative may not hold attention
Listeners emphasized that Reels, TikTok, and YouTube Shorts often offer humor, sex appeal, information, or a narrative arc, whereas Vibes served surreal visuals with bouncy music. One listener called it another step toward “atomized, isolated experiences” built to monetize attention.
Thompson accepted the criticism through his own publishing history. A technically strong analyst he once followed failed to attract subscribers because the work was dry; Stratechery succeeded partly because its best pieces supplied protagonists, beginnings, middles, and endings. “The lack of narrative or storytelling is not compelling to most people.”
His taste is genuinely atypical: ambient futuristic landscapes and Apple TV aerial screensavers feed an internal narrative already running “24/7,” while externally imposed stories can feel tiring. Vibes therefore resembled touching grass to him—“I like artificial grass. What can I say?”—but that preference created “a massive blind spot.”
Sharp admitted his own failure to push back: Thompson had noticed during their demo that “Andrew’s bored,” yet Sharp joined the enthusiastic endorsement. He now views Vibes and Sora as 1993–94-era computers—half-baked experiments whose present awkwardness may look charming once the next four or five years clarify the destination.
6. The broader criticism is attention decay
- The closing outside critique came from Ashkay Kothari: why devote “our brightest minds, billions of dollars, and the most powerful GPUs on Earth” to another app that “optimizes for attention decay”? Sharp said versions of that question had recurred every 10 minutes on Twitter for the prior seven days, aimed at both Meta Vibes and Sora. It was presented as a recurring audience concern, not a settled conclusion from either host.