内容来源：https://techcrunch.com/2025/12/15/vcs-discuss-why-most-consumer-ai-startups-still-lack-staying-power/

内容总结：

生成式人工智能浪潮兴起已三年，但多数AI初创公司仍主要面向企业而非个人消费者盈利。尽管ChatGPT等通用大语言模型迅速被公众接纳，但垂直领域的消费级AI应用尚未形成广泛吸引力。

风投机构Goodwater Capital管理合伙人简池华指出，早期围绕音视频、图像生成的AI应用曾引发热潮，但随着Sora、Nano Banana等模型及中国开源视频模型的涌现，许多细分市场机会已逐渐消失。他将这类应用比作2008年iPhone问世后昙花一现的第三方手电筒应用——当核心功能被操作系统原生集成，独立产品的生存空间便急剧收缩。

简池华认为，正如智能手机生态历经数年沉淀才孕育出Uber、Airbnb等变革性应用，AI平台同样需要“稳定期”才能催生持久的消费级产品。他表示：“我们正处在类似2009-2010年移动互联网爆发的前夜。”当前谷歌Gemini与ChatGPT达到技术 parity 的态势，或许正预示着生态渐趋成熟。

Scribble Ventures创始合伙人伊丽莎白·威尔将当前消费级AI应用的发展阶段形容为“尴尬的青春期”。两位投资者一致认为，智能手机可能并非AI消费产品的终极载体——其有限的感知能力（仅能捕捉用户3%-5%的视野）与非沉浸式交互模式，制约了AI潜力的充分发挥。

行业正积极探索智能手机的替代方案：OpenAI与前苹果设计总监乔尼·艾夫据传正在开发无屏口袋设备；Meta的雷朋智能眼镜通过腕带识别微手势操控；多家初创公司尝试开发AI胸针、吊坠或戒指等新型可穿戴设备，但市场反响尚不理想。

不过，并非所有AI消费产品都依赖硬件革新。简池华预测，定制化个人AI理财顾问可能成为突破口；威尔则预见“全天候个性化AI导师”将通过智能手机普及。但二人对近期涌现的AI社交网络创业潮持保留态度，简池华指出：“当数千个AI机器人围绕用户内容互动时，社交就变成了单人游戏。人们使用社交网络的核心动力，在于感知屏幕另一端真实存在的人类。”

（注：文末会议推广内容已根据要求省略）

中文翻译：

生成式人工智能热潮兴起已三年有余，但绝大多数AI初创企业仍主要面向企业客户而非个人消费者实现盈利。尽管ChatGPT等通用大语言模型迅速获得消费者青睐，但多数垂直领域的消费级生成式AI应用尚未引发市场共鸣。

"早期围绕视频、音频和照片的AI应用曾令人惊艳，"Goodwater Capital联合创始人兼管理合伙人简池华在十二月初的TechCrunch StrictlyVC活动舞台上表示，"但随着Sora和Nano Banana的问世，加上中国开源视频模型的推出，许多商业机会已然消失。"

他将这类应用比作简易手电筒应用——2008年iPhone问世后曾风靡一时的第三方应用，很快就被整合进iOS系统。简池华指出，正如智能手机平台需要数年沉淀才催生颠覆性消费应用，AI平台同样需要经历"稳定期"才能孕育持久的消费级产品。

"我认为我们正处在类似2009-2010年移动互联网爆发的前夜，"简池华表示。那个时期正是优步、爱彼迎等移动优先消费级巨头诞生的年代。他指出，随着谷歌Gemini达到与ChatGPT相当的技术水平，我们或许已能窥见这种稳定期的端倪。

Scribble Ventures创始人兼合伙人伊丽莎白·威尔同样认为，当前消费级AI应用正处于"尴尬的青春期过渡阶段"。消费级AI初创企业如何走向成熟？或许需要超越智能手机的新设备载体。

"智能手机每天被拿起500次，却只能感知你视野的3%到5%，这样的设备难以充分释放AI潜力，"简池华坦言。威尔赞同智能手机在很大程度上限制了消费级AI产品的重塑空间，主因其缺乏环境感知能力。她向观众展示自己的iPhone时说："我认为五年后我们不会再为此设备开发应用。"

初创企业与科技巨头正竞相研发取代智能手机的新型个人设备。OpenAI与苹果前设计总监乔尼·艾夫据传正在开发"无屏幕"口袋设备；Meta雷朋智能眼镜通过感知细微手势的腕带操控；多家初创企业尝试推出AI胸针、吊坠或戒指，但效果多不尽如人意。

不过并非所有AI消费产品都依赖新硬件。简池华认为，定制化个人AI理财顾问可能成为突破口；威尔则预见个性化"全天候"AI导师将通过智能手机普及。尽管看好AI潜力，两位投资人对新兴的AI社交网络初创企业持保留态度。简池华指出这类平台让数千AI机器人参与用户内容互动，"将社交变成单人游戏，这背离了社交网络连接真实人类的本质"。

英文来源：

Even three years after the generative AI boom started, most AI startups are still making money by selling to businesses, not individual consumers.
Although consumers quickly adopted general-purpose LLMs like ChatGPT, most specialized consumer GenAI applications have yet to resonate.
“A lot of early AI applications around video, audio, and photo were super cool,” said Chi-Hua Chien, co-founder and managing partner at Goodwater Capital, onstage at TechCrunch’s StrictlyVC event in early December. “But then Sora and Nano Banana came out, and the Chinese open sourced their video models. And so, a lot of those opportunities disappeared.”
Chien compares some of those applications to the simple flashlight, which was initially a popular third-party download after the iPhone launched in 2008 but was quickly integrated into iOS itself.
He argued that, just as it took a few years for the smartphone platform to solidify before game-changing consumer apps emerged, AI platforms need a similar period of “stabilization” for lasting AI consumer products to flourish.
“I think we’re right on the cusp of the equivalent to mobile of the 2009-2010 era,” Chien said. That period was the birth of massive mobile-first consumer businesses like Uber and Airbnb.
We could be seeing inklings of that stabilization with Google’s Gemini reaching technological parity with ChatGPT, Chien said.
Join the Disrupt 2026 Waitlist
Add yourself to the Disrupt 2026 waitlist to be first in line when Early Bird tickets drop. Past Disrupts have brought Google Cloud, Netflix, Microsoft, Box, Phia, a16z, ElevenLabs, Wayve, Hugging Face, Elad Gil, and Vinod Khosla to the stages — part of 250+ industry leaders driving 200+ sessions built to fuel your growth and sharpen your edge. Plus, meet the hundreds of startups innovating across every sector.
Join the Disrupt 2026 Waitlist
Add yourself to the Disrupt 2026 waitlist to be first in line when Early Bird tickets drop. Past Disrupts have brought Google Cloud, Netflix, Microsoft, Box, Phia, a16z, ElevenLabs, Wayve, Hugging Face, Elad Gil, and Vinod Khosla to the stages — part of 250+ industry leaders driving 200+ sessions built to fuel your growth and sharpen your edge. Plus, meet the hundreds of startups innovating across every sector.
Elizabeth Weil, founder and partner at Scribble Ventures, echoed Chien’s sentiment about the early days of GenAI, describing the current state of consumer AI applications as being in an “awkward teenage middle ground.”
What will it take for consumer AI startups to grow up? Possibly a new device beyond the smartphone.
“It’s unlikely that a device that you pick up 500 times a day but only sees 3% to 5% of what you see is going to be what ultimately introduces the use cases that take full advantage of AI’s capabilities,” Chien said.
Weil agreed that a smartphone may be too limiting for reimagining consumer AI products in large part because it is not ambient. “I don’t think we’re going to be building for this in five years,” she said, indicating her iPhone as she showed it to the audience.
Startups and incumbent tech companies have been racing to build a new personal device that can supplant smartphones.
OpenAI and Apple’s former design chief, Jonny Ive, are working on what’s rumored to be a “screenless,” pocket-sized device. Meta’s Ray-Ban smart glasses are controlled by a wristband that detects subtle gestures. Meanwhile, a number of startups are trying, with often disappointing results, to introduce a pin, pendant, or ring that uses AI in a way different from how smartphones do.
However, not every AI consumer product will be dependent on a new device. Chien suggested that one such offering could be a personal AI financial adviser customized to the user’s specific needs. Similarly, Weil anticipates that a personalized, “always-on” tutor will become ubiquitous, with its specialized tutelage delivered directly from a smartphone.
Though excited by AI’s potential, Weil and Chien expressed skepticism about the emergence of several, still-stealthy AI-powered social network startups. Chien said these companies are building networks where thousands of AI bots are interacting with the user’s content.
“It turns social into a single-player game. I’m not sure that it works,” he said. “The reason that people enjoy social networking is the understanding that there are real humans on the other side.”