← 返回索引

报告 #13: 高付费-Mark

AI付费用户态度与决策行为研究

受访者:Mark
年龄:45
职业:Client Support Specialist
收入:< $30,000
AI工具:ChatGPT, Gemini, Midjourney, Copilot, DALL-E, Sora
来源:trooly.ai

📊 分析摘要

45岁男性,白人,Client Support Specialist,< $30,000。 使用:ChatGPT, Gemini, Midjourney, Copilot。 付费:$25-$400/月。 特征:考虑取消,价格敏感。 "Well, lately, it's usually just signing up people for the free student benefits package. - 噢,最近通常只是..."

💡 商业洞见(AI软件销售视角)

Background

Participant: Mark

受访者: Mark

Gender: Man

性别: 男

Age: 45

年龄: 45岁

Location: Milan, Indiana, United States

所在地: 美国印第安纳州米兰

Household Income: < $30,000

家庭年收入: < $30,000

Level of Education: Undergraduate degree

教育程度: 本科学位

Ethnicity: White

种族: 白人

Occupation: Client Support Specialist

职业: 客户支持专员

Industry: Cloud Software / SaaS

行业: 云软件 / SaaS

Children: No children

子女情况: 无子女

📜 查看完整访谈记录(283 段对话)

00:03-00:22 - Researcher: Hey there! Thanks so much for joining me today. I'm Trooly, and I'll be chatting with you for about 60 minutes to learn more about your experiences and perspectives. Before we get started, could you please give me a brief self-introduction? I'd love to get to know you a bit better. For example, you could tell me your age, occupation, and hobbies — or anything else you think would help me understand you better. - 嘿,你好!非常感谢你今天能参加访谈。我是 Trooly,接下来的 60 分钟左右我将与你聊天,以了解更多关于你的经历和见解。 在开始之前,你能简单做一个自我介绍吗?我想对你有更多了解。例如,你可以告诉我你的年龄、职业和爱好——或者任何你认为能帮我更好了解你的信息。

00:35-00:38 - Mark: Well, um, I'm forty five. Are you listening to me? Forty five years old. Work online. - 嗯,那个,我四十五岁。你在听吗?四十五岁。在网上工作。

00:41-00:46 - Researcher: Yes, I'm listening! Forty-five, and you work online. What kind of online work do you do? - 是的,我在听!四十五岁,在网上工作。 你做的是哪种在线工作?

00:48-00:52 - Mark: I answer email for a service as a software company. - 我在一家软件公司为一项服务回复邮件。

00:54-01:00 - Researcher: Got it, so you handle email support for a software company. Where are you based, and what do you like to do outside of work? - 明白了,所以你是在一家软件公司负责邮件支持。 你住在哪里,工作之余喜欢做些什么?

01:22-01:27 - Mark: I like listening to music and stuff pretty much. Like, just hanging out, watching TV, - 我挺喜欢听音乐之类的。比如,就是到处闲逛,看电视。

01:29-01:30 - Researcher: Nice, pretty - 挺好的,挺……

01:30-01:31 - Mark: Are you listening still? - 你还在听吗?

01:33-01:43 - Researcher: Yes, absolutely! I'm here and listening to everything you're saying. You mentioned music and TV - what kind of music are you into? - 是的,当然在听!我在这儿,正听着你说的一切呢。 你提到了音乐和电视——你喜欢哪种类型的音乐?

01:45-01:55 - Mark: Well, I like Nirvana. Tool. Deftones, Bush, certainly, like, all the grunge things from the nineties. - 嗯,我喜欢 Nirvana(涅槃乐队)。Tool。Deftones,还有 Bush,当然,还有所有九十年代的垃圾摇滚。

01:57-02:08 - Researcher: Cool, so you're into the 90s grunge scene - Nirvana, Tool, Deftones, Bush. That's a solid lineup. Tell me more about your work handling emails for the software company. What does a typical workday look like for you? - 酷,所以你喜欢 90 年代的垃圾摇滚——Nirvana、Tool、Deftones、Bush。这阵容很扎实。 多跟我说说你在软件公司处理邮件的工作吧。你典型的工作日是怎样的?

02:17-02:26 - Mark: Well, I get up in the morning, get something to drink, wake up with some cold ice water, and then try to jump in the shower before I get logged on to the Internet. - 嗯,我早上起床,喝点东西,用凉冰水让自己清醒一下,然后在登录互联网之前试着冲个澡。

02:28-02:38 - Researcher: So you start with water, shower, then log on. Once you're online, walk me through what happens - like, what kinds of emails are you dealing with, and how does your day flow from there? - 所以你先喝水,洗澡,然后登录。 一旦你上线了,请带我了解一下接下来的流程——比如,你处理的是什么类型的邮件,你的一天是如何展开的?

02:59-03:11 - Mark: Oh, well, usually, it's just a file conversion website. So, usually, it's a bunch of students currently who are trying to convert their documents, like, PDF files to Microsoft Word documents, for example, and they have to use our website to turn it in the correct format for their teachers. - 哦,好吧,通常这只是一个文件转换网站。所以,通常目前是一群学生试图转换他们的文档,例如,把 PDF 文件转换成 Microsoft Word 文档,他们必须使用我们的网站将其转换成正确的格式交给老师。

03:13-03:21 - Researcher: Okay, so students converting files - PDFs to Word and things like that. What kinds of issues do they usually email about? Like, what goes wrong that they need your help? - 明白了,所以是学生转换文件——PDF 转 Word 之类的。 他们通常会为了什么问题发邮件?比如,出了什么问题需要你的帮助?

03:31-03:33 - Mark: Well, lately, it's usually just signing up people for the free student benefits package. - 噢,最近通常只是帮人们注册免费的学生福利包。

03:36-03:44 - Researcher: Interesting - so there's a free student benefits package they can sign up for. What does that package include, and why do students need help signing up for it? - 有意思——所以有一个他们可以注册的免费学生福利包。 那个福利包包含什么,为什么学生在注册时需要帮助?

03:58-04:05 - Mark: Oh, well, usually, it's automatic, but sometimes we get emails from weird schools or unorthodox that aren't from the EDU domain, so we gotta take a look at each one of them. - 哦,通常它是自动的,但有时我们会收到来自奇怪学校或非正规学校的邮件,那些邮箱不是 .edu 后缀的,所以我们得逐个查看。

04:08-04:16 - Researcher: Ah, so most .edu emails go through automatically, but the non-standard schools need manual review. How many of these emails are you typically handling in a day? - 啊,所以大多数 .edu 邮件会自动通过,但非标准的学校需要人工审核。 你通常一天要处理多少封这类邮件?

04:17-04:23 - Mark: Maybe one hundred on a decent day. - 状态还不错的一天大概有一百封吧。

04:26-04:34 - Researcher: A hundred emails on a decent day - that's quite a bit. How long have you been doing this work, and are you working for yourself or as part of a team? - 状态不错的一天处理一百封邮件——那挺多的。 你做这份工作多久了,你是为自己工作还是团队的一员?

04:41-04:44 - Mark: I've been doing this exact role for about ten years, and it's just myself. - 我做这个岗位大概十年了,就我一个人。

04:46-04:55 - Researcher: Ten years doing this solo - wow. If you don't mind me asking, what's your monthly income from this work? Just a rough range is fine, or if you'd rather not say, that's totally okay too. - 独自做了十年——哇。 如果你不介意我问的话,你这份工作的月收入是多少?大概的范围就可以,如果你不想说也没关系。

04:59-05:00 - Mark: From this job, - 就这份工作而言,

05:01-05:02 - Researcher: Yeah, - 是的,

05:03-05:03 - Mark: I only get about four hundred a month. - 我每个月只拿到四百美元左右。

05:07-05:12 - Researcher: Four hundred a month from this job. Do you have any other income sources besides the email support work? - 这份工作每个月四百美元。 除了这份邮件支持工作,你还有其他收入来源吗?

05:13-05:15 - Mark: Not at this time. - 目前没有。

05:19-05:29 - Researcher: Got it, so $400 a month is your main income right now. Now let's shift gears and talk about AI. What AI apps or web tools do you use most often in your life? Tell me a bit about your experience with them. - 明白了,所以每月 400 美元是你目前的主要收入。 现在让我们换个话题来聊聊 AI。你在生活中最常使用哪些 AI 应用或网页工具?跟我分享一下你使用它们的经验。

05:44-05:56 - Mark: I think I've tried all of them. I've tried ChatGPT. I started with Copilot on Microsoft. I don't really use that anymore. ChatGPT was big. Gemini was big. Back when it was Bard. I think I've seen all of them come up Right now, I'm using Meta AI which is based on Llama. - 我觉得我所有的都试过了。我试过 ChatGPT。我是从微软的 Copilot 开始的。我现在不怎么用那个了。ChatGPT 曾经动静很大。Gemini 也很火。那是它还叫 Bard 的时候。我觉得我见证了它们所有的兴起。现在,我正在使用基于 Llama 的 Meta AI。

05:59-06:09 - Researcher: So you've been through the whole lineup - Copilot, ChatGPT, Gemini (back when it was Bard), and now you're mainly using Meta AI. What made you switch to Meta AI? What do you like about it compared to the others? - 所以你把全家桶都用遍了——Copilot、ChatGPT、Gemini(在它还叫 Bard 的时候),现在主要使用 Meta AI。 是什么让你转向 Meta AI 的?比起其他的,你喜欢它哪一点?

06:21-06:30 - Mark: I guess well, because I got these Meta glasses. So it it goes right with their app or whatever. - 我猜是因为我买了这些 Meta 眼镜。所以它直接匹配它们的 App 之类的。

06:32-06:43 - Researcher: Ah, you have Meta glasses, so it integrates directly with them. Are you currently paying for any of these AI tools - ChatGPT, Meta AI, or any others? Tell me about which ones you're paying for and how much. - 啊,你有 Meta 眼镜,所以它直接集成了。 你目前有在为这些 AI 工具付费吗——ChatGPT、Meta AI 或者其他的?告诉我你正在为哪些付费,以及付了多少钱。

06:55-07:02 - Mark: Well, I guess, I'm paying for the chat GPT. But that's - 嗯,我猜,我在付 ChatGPT 的钱。但那是……

07:07-07:14 - Researcher: Okay, you're paying for ChatGPT. How much are you paying per month, and which plan do you have - is it the Plus plan or something else? - 好的,你在为 ChatGPT 付费。 你每个月付多少钱,用的是什么计划——是 Plus 计划还是别的?

07:15-07:26 - Mark: Yeah. It's the plus plan for sure. - 是的。肯定就是 Plus 计划。

07:28-07:32 - Researcher: The Plus plan - so that's $20 a month, right? Are you paying for any - Plus 计划——所以每个月是 20 美元,对吗? 你还在为……付费吗?

07:36-07:41 - Mark: Yeah. Well, then it then the meta glasses I'm just paying for them as a subscription. - 对。然后那个 Meta 眼镜,我只是以订阅的方式付费。

07:41-07:49 - Researcher: Are you Wait, so the Meta glasses themselves are on a subscription? How does that work - what are you paying monthly - 等一下,所以 Meta 眼镜本身是订阅制的?它是怎么运作的——你每个月付多少……

07:49-07:56 - Mark: Oh, no. Oh, no. Sorry. I mean, the verified badge on meta. - 哦,不。哦,不。抱歉。我是指 Meta 上的认证徽章(Meta Verified)。

07:58-08:06 - Researcher: Ah okay, so you're paying for Meta Verified - the blue badge subscription. How much is that per month, and what made you decide to get the verified badge? - 啊好的,所以你在付 Meta Verified 的钱——就是那个蓝标订阅。 那个每个月多少钱,是什么让你决定购买认证徽章的?

08:16-08:18 - Mark: It's twelve dollars, and I was just getting that so I could try out all the features of the new glasses. I apologize for my mistake. - 十二美元,我买那个是为了试用新眼镜的所有功能。为我的错误表达道歉。

08:20-08:33 - Researcher: No worries at all! So $12 for Meta Verified to access all the features with the glasses. Besides ChatGPT Plus and Meta Verified, are you paying for any other AI tools - maybe for content creation, image generation, or work-related - 完全没关系!所以是为了配合眼镜使用所有功能,付了 12 美元的 Meta Verified。 除了 ChatGPT Plus 和 Meta Verified,你还在为其他 AI 工具付费吗——比如用于内容创作、图像生成或工作相关的……

08:33-08:34 - Mark: Yes. Or - 是的。或者……

08:38-08:39 - Researcher: What else are you paying for? - 你还在为什么付钱?

08:42-08:43 - Mark: Sora. You know that. - Sora。你知道那个吧。

08:46-08:47 - Researcher: Sora - the OpenAI - Sora——就是 OpenAI 的……

08:47-08:47 - Mark: Or you know Sora? Yeah. - 你知道 Sora?是的。

08:50-08:58 - Researcher: Yeah, the OpenAI video generation tool. How much are you paying for Sora, and what's the payment model - is it a monthly subscription or pay-per-use? - 是的,OpenAI 的视频生成工具。 你为 Sora 付多少钱,付费模式是怎样的——是月度订阅还是按次付费?

09:00-09:03 - Mark: It's just it's bundled with ChatGPT. - 它就是和 ChatGPT 捆绑在一起的。

09:06-09:16 - Researcher: Oh got it, so Sora access comes bundled with your ChatGPT Plus subscription - you're not paying separately for it. Any other AI tools you're paying for besides ChatGPT Plus and Meta Verified? - 哦明白了,所以 Sora 的访问权限是包含在你的 ChatGPT Plus 订阅里的——你并没有额外付钱。 除了 ChatGPT Plus 和 Meta Verified,还有其他你在付费的 AI 工具吗?

09:20-09:26 - Mark: Nope. That's it for now. I used to have a free trial of a bunch of different ones. - 没了。目前就这些。我以前试过一大堆不同工具的免费试用。

09:29-09:42 - Researcher: Okay, so right now it's just ChatGPT Plus at $20 and Meta Verified at $12 - that's $32 a month total. You mentioned free trials of a bunch of different ones - which ones did you try, and why didn't you continue paying for them after the trial ended? - 好的,所以现在就是 20 美元的 ChatGPT Plus 和 12 美元的 Meta Verified——总共每月 32 美元。 你提到过试用了一大堆不同的工具——你试过哪些,为什么试用结束后没有继续付费?

09:46-10:03 - Mark: Well, for one, I'll tell you the Microsoft Copilot came with Windows, and they kinda, like, forced it on to you. So I didn't I I don't use Microsoft Windows whatsoever anymore, so I don't I haven't used Copilot since I haven't used Windows. - 噢,首先,我告诉你,微软的 Copilot 是随 Windows 附带的,他们有点像强加给你。所以我已经——我完全不再使用微软 Windows 了,所以自从不用 Windows,我也就没用过 Copilot 了。

10:05-10:13 - Researcher: So you stopped using Copilot when you stopped using Windows altogether. What other free trials did you try, and what made you decide not to keep paying for those? - 所以当你彻底停用 Windows 时,你也停止使用 Copilot 了。 你还试过哪些免费试用,是什么让你决定不继续付费的?

10:15-10:17 - Mark: Let me think. I've tried Midjourney before. - 让我想想。我以前试过 Midjourney。

10:20-10:27 - Researcher: Mid Journey - the image generation tool. What did you think of it during the trial, and why didn't you keep the subscription going? - Midjourney——图像生成工具。 试用期间你觉得它怎么样,为什么没有继续订阅?

10:28-10:44 - Mark: Think it was great but then I was using so many, and it was kind of a big thing. Everybody had one. So I was trying like, DALL-E and Gemini was coming out - 我觉得它很棒,但当时我用的太多了,而且这事儿挺大的。每个人都有一个。所以我当时在试用 DALL-E,Gemini 也要出来了……

10:47-10:59 - Researcher: So there were just too many at once and you were experimenting with all of them - Midjourney, DALL-E, Gemini. When you decided which ones to actually pay for long-term, what made ChatGPT the one you stuck with instead of the others? - 所以当时一下子有太多工具了,你都在尝试——Midjourney、DALL-E、Gemini。 当你决定长期为哪些工具付费时,是什么让 ChatGPT 成为你坚持使用、而非其他的那个?

11:12-11:14 - Mark: Well, I guess I just kind of they they're the ones that that tried and true stuck around, the ones I always went back to. - 嗯,我猜是因为它们是那些经受住考验并保留下来的,我总是会回过头去用的那一个。

11:18-11:25 - Researcher: They're the ones you kept coming back to. What specifically do you use ChatGPT for that keeps you coming back? What are your most common use cases? - 它们是你一直回过头去使用的。 你具体用 ChatGPT 做什么让你离不开它?你最常用的场景是什么?

11:30-11:34 - Mark: Well, mainly, I use it for just everyday facts and stuff, but sometimes I use it for image generation. - 嗯,主要就是查一些日常事实之类的,但有时我也用它生成图像。

11:36-11:45 - Researcher: Everyday facts and image generation. When you say everyday facts - can you give me a specific example from the last week or two? Like, what's something you recently asked it? - 日常事实和图像生成。 当你说“日常事实”时——你能给我一个过去一两周里的具体例子吗?比如,你最近问了它什么?

12:24-12:27 - Mark: Trying to get the the thing pulled up, but I can't. - 我正试着把那个记录调出来,但在我弄不出来。

12:28-12:37 - Researcher: That's okay, no need to pull it up right now. Just from memory - what's a typical question you might ask it? Like, is it more research stuff, help with decisions, or something else? - 没关系,现在不需要调出来。 凭记忆说说看——一个典型的问题可能是怎样的?比如,更多是研究类的内容、辅助决策,还是其他的?

12:40-12:47 - Mark: Oh, well, everyday things like, how's the weather? But also just certain things like when I'm cooking about measurements. - 噢,比如日常的事情,像天气怎么样?但也包括一些特定的事,比如我做饭时的计量转换。

12:49-13:00 - Researcher: Got it - quick practical stuff like weather and cooking conversions. Recently, has there been any AI purchase or subscription that made you feel really satisfied - like you felt it was totally worth the money? - 明白了——快速实用的东西,比如天气和烹饪转换。 最近,有没有什么 AI 购买或订阅让你感到非常满意——让你觉得完全物有所值?

13:02-13:06 - Mark: ChatGPT was doing pretty good, so I'll I'll stick with that. - ChatGPT 表现一直挺不错,所以我会坚持用它。

13:08-13:18 - Researcher: ChatGPT's been doing pretty good for you. What about it specifically makes you feel like it's worth the $20 a month? What's the thing that makes you think "yeah, this is worth paying for"? - ChatGPT 对你来说一直挺好用的。 具体是什么让你觉得它值得每月 20 美元?哪一点让你觉得“是的,这钱花得值”?

13:26-13:29 - Mark: Well, it's gotta be fast. It has to have like, relevant feedback. Can't crash all the time. - 唔,它必须得快。它必须有相关的反馈。不能总是崩溃。

13:33-13:44 - Researcher: Speed, relevant answers, and reliability - can't be crashing. Have you had any AI purchases recently that really disappointed you - where you felt like it wasn't worth the money or didn't live up to what you expected? - 速度、相关的回答以及可靠性——不能总是崩。 最近有没有什么 AI 购买让你非常失望——你觉得不值这个钱或者没达到你的预期?

13:50-14:10 - Mark: Recently, it's just the image generation on Meta AI. Has been kind of weird, but I I guess I didn't pay for it. So I got the professional to see if the image generation quality changes and it doesn't really. I think they might just remove the watermark. - 最近就是 Meta AI 的图像生成。一直有点奇怪,不过我猜我没付钱。所以我买了专业版(认证版)看看图像生成的质量是否会改变,结果其实并没变。我觉得他们可能只是去掉了水印。

14:14-14:27 - Researcher: So you upgraded to Meta Verified hoping the image generation would improve, but it's still kind of weird - seems like they just removed the watermark and that's it. What's "weird" about the image generation? What's not working the way you want? - 所以你升级到 Meta Verified 是希望能提升图像生成质量,但它还是有点奇怪——似乎他们只是去掉了水印,仅此而已。 图像生成哪里“奇怪”?哪里不符合你的要求?

14:30-14:44 - Mark: Oh, well, I just thought it would be more accurate or defined or refined or something. It is the it is the same image generator, though. So nothing really changed. - 噢,我只是以为它会更准确,或者定义更清晰,或者更精细之类的。但它确实是同一个图像生成器。所以真的没什么变化。

14:47-15:01 - Researcher: So the image quality didn't get more accurate or refined with the paid version - it's literally the same generator. Let's dive into AI content creation tools more specifically. For things like generating or editing images, videos, or audio - which AI tools do you actually use regularly? - 所以付费版的图像质量并没有变得更准确或更精细——它字面上就是同一个生成器。 让我们更深入地谈谈 AI 内容创作工具。对于像生成或编辑图像、视频或音频这类事情——你实际经常使用哪些 AI 工具?

15:09-15:25 - Mark: Yeah. Regularly, um, Meta AI, I can I usually just go in there and tell it to redo something, but now I'm finding that its app has video and a whole bunch of other things to compete with Sora? - 对。经常用的,嗯,Meta AI,我通常可以直接进去让它重做某样东西,但现在我发现它的 App 有视频和一大堆其他功能来和 Sora 竞争?

15:27-15:36 - Researcher: Interesting - so Meta AI now has video generation features to compete with Sora. Besides Meta AI, what other AI content creation tools are you using regularly for images, video, or audio? - 有意思——所以 Meta AI 现在有了视频生成功能,要和 Sora 竞争。 除了 Meta AI,你还在经常使用哪些 AI 内容创作工具来制作图像、视频或音频?

15:48-15:57 - Mark: I've had Gemini use uh, I've tried VisionLens before, but I have Gemini used to identify what's inside of a picture. It's way better than Meta's Llama. - 我用过 Gemini 的……呃,我以前试过 VisionLens,但我用 Gemini 来识别图片里的东西。它比 Meta 的 Llama 强多了。

15:59-16:12 - Researcher: So you use Gemini for image recognition - identifying what's in a picture - and it's way better than Meta's Llama at that. What about for actually creating or generating content - are you using ChatGPT's image generation, or any other tools besides Meta AI? - 所以你用 Gemini 进行图像识别——识别图片里的内容——在这方面它比 Meta 的 Llama 好得多。 那实际创建或生成内容呢——你有在使用 ChatGPT 的图像生成,或者 Meta AI 以外的其他工具吗?

16:20-16:32 - Mark: No. Uh, I've actually I made my company's logo in Copilot, the new logo. So I used I actually used Copilot more for image generation, but I no longer use that at all. - 没有。呃,我其实在 Copilot 里做了我公司的 Logo,新版 Logo。所以我其实更多用 Copilot 生成图像,但我现在一点也不用它了。

16:35-16:49 - Researcher: You made your company's logo in Copilot, but you don't use it anymore since you stopped using Windows. So right now, for creating images, you're mainly using Meta AI. Can you walk me through two recent examples - what prompts did you give it and what were you trying to create? - 你在 Copilot 中制作了公司 Logo,但自从不用 Windows 后你就不再用它了。 所以现在,为了创建图像,你主要使用 Meta AI。你能带我回顾两个最近的例子吗——你给了它什么提示词,你想创建什么?

17:04-17:05 - Mark: Yeah. Recently, I just had this little picture that was from my nephew's drew of, like, two monsters fighting, and I told the AI to turn it into realistic monsters. - 好的。最近我有一张我侄子画的小图,画的是两个怪物在战斗,我就告诉 AI 把这画变成真实的怪物。

17:08-17:18 - Researcher: That's cool - you had your nephew's drawing of two monsters and asked Meta AI to turn it into realistic monsters. What was the second recent example? What else have you generated with it lately? - 太酷了——你拿着你侄子画的两个怪物的画,让 Meta AI 把它变成真实的怪物。 第二个最近的例子是什么?你最近还用它生成了什么?

17:29-17:32 - Mark: Used the AI to, like, as an inpainting tool to change the text of a logo on someone's T shirt in an image. - 用 AI 作为局部重绘(inpaint)工具,在一张照片里修改某人 T 恤上的 Logo 文字。

17:35-17:47 - Researcher: So you used it as an inpainting tool to change text on a logo in a photo of someone's t-shirt. For both of these - the realistic monsters and the t-shirt edit - what was the purpose? Were these for fun, for work, or something else? - 所以你把它当成局部重绘工具,修改了一个穿着 T 恤的人照片里的 Logo 文字。 这两件事——真实怪物和 T 恤编辑——目的是什么?是为了玩、为了工作,还是别的?

17:51-18:25 - Mark: Yeah. Well, the Monsters was just for fun. See, AI has become where it started out as just for fun, then I used it for work, and now it is just part of life. So it's that's something I just did in in five minutes. Let's make AI out of it. Whereas before, it wasn't even an option. The other thing was yeah. It was just for fun. It was for a meme to just change what someone's shirt actually said to take their text off, put some new text on, and it's just instant change. So five minutes - 噢。那个怪物纯粹是为了玩。你看,AI 经历了一个过程,刚开始只是为了玩,然后我把它用于工作,而现在它只是生活的一部分了。所以那只是我花五分钟做的事。比如“让我用 AI 做一个吧”。而在以前,这甚至都不是一个选项。另一件事也是……是的,只是为了玩。是为了做一个梗图(meme),通过改变某人衣服上实际的内容,把他们的文字删掉,放上一些新文字,而且这种改变是瞬间的。所以就五分钟。

18:27-18:42 - Researcher: So AI has shifted from "just for fun" to "for work" to now just being part of everyday life - something you do in five minutes without thinking about it. When did you first start using Meta AI for image generation? Do you remember where you first heard about it or what made you try it? - 所以 AI 已经从“只是为了玩”转变为“为了工作”,现在变成了日常生活的一部分——一种你花五分钟不假思索就能做的事。 你最早是什么时候开始使用 Meta AI 进行图像生成的?你记得是在哪里第一次听说它,或者是什么让你尝试的?

18:47-18:58 - Mark: I think it popped up in our group chat, which we're on pretty much every day. So that's where it started. I'm sure it had to it had to happen with a suggestion from Meta. - 我觉得它是在我们的群聊里弹出来的,我们几乎每天都在那个群里。那就是开始的地方。我敢肯定这与 Meta 官方的推荐有关。

19:01-19:12 - Researcher: It popped up in your group chat - probably Meta suggesting it - and since you're in that chat every day, you just started using it. How long have you been using Meta AI for image generation? Like, was this months ago, or more recent? - 它在你的群聊里弹出来——可能是 Meta 推荐的——既然你每天都在那个群里,你就开始用了。 你使用 Meta AI 生成图像多久了?比如,是几个月前,还是最近的事?

19:16-19:35 - Mark: Years. I know. I I guess when did it start coming in here? It's all kind of blurred together since this pandemic. But I think they let me have it early adopter maybe even. It's had to been at least a year or two. - 好几年了。我知道……我猜它是什么时候开始进入这里的?自从这次大流行之后,一切都变得模糊了。但我想他们甚至让我成为了早期用户。肯定至少有一两年了。

19:37-19:49 - Researcher: At least a year or two, maybe as an early adopter - it's all blurred together since the pandemic. Now let's talk about Sora. You mentioned you're paying for it through your ChatGPT subscription. What are you actually using Sora for? - 至少有一两年了,可能是作为早期用户——大流行以来,时间感都变模糊了。 现在让我们聊聊 Sora。你提到你是通过 ChatGPT 订阅付费获取它的。你实际上用 Sora 做什么?

20:12-20:17 - Mark: I just wanted to see it was, the the leading video generator. So I had to test it out. I had to check it out. It requires, like, you have to have an Apple device and some special thing. And then I found out the meta app does kinda the same thing, and I'm already, like, a meta guy and not really an Apple guy. So we're sticking with this. - 我只是想看看它是否是领先的视频生成器。所以我必须测试一下。我得去查查。它要求……比如你得有一个苹果设备和一些特殊的东西。然后我发现 Meta 的 App 也能做差不多的事,而且我已经是个 Meta 粉了,算不上苹果粉。所以我就坚持用这个了。

20:19-20:35 - Researcher: So you wanted to test Sora because it was the leading video generator, but it requires Apple devices and special setup. Then you found out Meta's app does similar video generation, and you're already a Meta person, not an Apple person. Have you actually created any videos with Sora yet, or is it more just having access to try it? - 所以你想测试 Sora 是因为它是领先的视频生成器,但它需要苹果设备和特殊的设置。后来你发现 Meta 的应用也有类似的视频生成功能,而你已经是 Meta 的忠实用户,不是苹果用户。 那你目前实际上用 Sora 创建过任何视频了吗,还是说只是拥有去尝试的权限?

20:48-20:51 - Mark: It was just having access to try it because it was, like, an exclusive little club. People were making things, and then they were getting out to other platforms. And leaking out. And that's how I found out about it. - 只是拥有试用的权限,因为那时候它就像个排外的小俱乐部。人们在制作东西,然后东西流传到其他平台。泄露了出来。我就是这样发现它的。

20:55-21:07 - Researcher: So it felt like an exclusive club - people were creating things and leaking them to other platforms, and that's how you found out about it. Now that you've had access for a bit, have you tried generating any videos with it? What did you make? - 所以感觉像个专属俱乐部——人们创作东西并泄露到其他平台,你就是这样知道它的。 既然你已经获得权限有一段时间了,你试过用它生成视频吗?你做了什么?

21:09-21:18 - Mark: Yeah. Oh, I made, like, a bunch of obviously cat videos, like flying cats and rocket cats and spaceship cats. That's about it. - 试过。噢,我做了一堆……显然是猫的视频,比如飞天猫、火箭猫和太空船猫。大概就这样。

21:19-21:28 - Researcher: Flying cats, rocket cats, spaceship cats - classic. What about with Meta's video generation that you mentioned? Have you tried making anything with that yet, or is it too new? - 飞天猫、火箭猫、太空船猫——经典。 那你提到的 Meta 的视频生成呢?你试过用那个做任何东西吗,还是它太新了?

21:37-21:41 - Mark: Yeah. That one was funny. I made myself walking like a giant ice queen out of a Disney movie or something. It was funny. - 试过。那个挺逗的。我把自己做成了一个像迪斯尼电影里的巨大冰雪女王那样在行走。挺搞笑的。

21:44-21:55 - Researcher: Ha, you as a giant ice queen walking out of a Disney movie - that's pretty funny. Between Sora and Meta's video generation, which one are you finding yourself using more, or are they about the same quality? - 哈,把你做成一个像迪斯尼电影里走出来的巨大冰雪女王——那确实挺有意思。 在 Sora 和 Meta 的视频生成之间,你发现自己更多在用哪一个,或者它们的质量差不多?

22:04-22:04 - Mark: Well, you know, since I'm all on board with Meta now, So I'm gonna stick with Meta. - 嗯,你知道的,既然我现在全身心支持 Meta,那我会坚持用 Meta。

22:08-22:20 - Researcher: You're all in on Meta, so you're sticking with their video generation. Given that - how likely are you to keep paying for Sora access through ChatGPT? On a scale of 0 to 10, where 0 is definitely canceling and 10 is definitely keeping it? - 你全面支持 Meta,所以你会坚持用他们的视频生成。 鉴于此——你有多大可能继续通过 ChatGPT 为 Sora 的访问权付费?按 0 到 10 分评分,0 是肯定取消,10 是肯定保留?

22:23-22:25 - Mark: Big ole zero. - 妥妥的 0 分。

22:27-22:36 - Researcher: Zero - so you're planning to drop it. Is that because Meta's doing the same thing for less money, or is there something else about Sora that's not working for you? - 0 分——所以你计划退订。是因为 Meta 花更少的钱就能做同样的事,还是 Sora 还有其他什么不符合你要求的地方? 是因为 Meta 花更少的钱就能做同样的事,还是 Sora 还有其他什么不符合你要求的地方?

22:42-22:42 - Mark: Yeah. It's true. Meta is doing the same thing for for free, actually. - 是的,没错。Meta 实际上是免费提供同样的功能。

22:45-22:58 - Researcher: Meta's doing the same thing for free - that's a pretty clear reason to drop it. When you think about your $20 ChatGPT Plus subscription that includes Sora, are you considering canceling the whole thing, or just wish you could drop the Sora part? - Meta 免费提供同样的功能——这确实是放弃它的一个非常充分的理由。 当你想到你那包含 Sora 的 20 美元 ChatGPT Plus 订阅时,你是考虑取消整个订阅,还是仅仅希望可以去掉 Sora 那部分?

23:01-23:05 - Mark: It was just to test out anyways. So just the Sora part. - 反正当初也只是为了测试。所以只是 Sora 那部分。

23:10-23:26 - Researcher: Got it - Sora was just to test out anyway, so you'd drop that part if you could. Let me ask about the image generation side: you're using Meta AI for images now. Before, you said you made your company logo in Copilot. When you were using Copilot for image generation, how did that compare to what Meta AI does now? - 明白了——Sora 反正只是为了测试,所以如果可以的话你会去掉那部分。 让我问问图像生成方面:你现在正在使用 Meta AI 生成图像。之前你提到你在 Copilot 里制作了公司的 Logo。当你使用 Copilot 生成图像时,它和现在的 Meta AI 相比如何?

23:38-23:50 - Mark: Well, Copilot was fast and it didn't even really have a watermark in the beginning. So it was pretty cool. Uh, I was comparing them all, and that was just the one that had the most gnarly radical as I call it. And it was a shadow tech logo, so it was outrageous. We loved it. - 嗯,Copilot 很快,而且一开始甚至都没有水印。所以非常酷。呃,我当时在对比所有的工具,那个就是结果最“炸裂”(gnarly radical,我这么称呼它)的一个。那是个影技术(shadow tech)Logo,所以效果非常惊人。我们很喜欢。

23:53-24:07 - Researcher: Copilot was fast, no watermark at first, and it gave you the most "gnarly radical" results - perfect for a shadow tech logo. What made you stop using Copilot for images once you switched away from Windows? Could you still access it another way, or was it just tied to the Windows experience? - Copilot 很快,起初没有水印,而且它给了你最“炸裂”的结果——非常适合影技术 Logo。 是什么让你在不再使用 Windows 后也停止使用 Copilot 处理图像?你还能通过其他方式访问它吗,还是它仅仅是和 Windows 的体验绑定在一起的?

24:19-24:36 - Mark: I think, actually, the reason I started to hate it was because they forced it on me in Windows. So now it's built into every new computer I think. We're all gonna have hardware that's used to it, and we're all caught up and used to it as humans now. A little bit better instead of it's just a different shift. It's a different time. So it'll be fine. - 我觉得,其实我开始讨厌它的原因是因为他们在 Windows 里强加给我。所以现在它内置在每一台新电脑里了。我想我们的硬件都会适应它,而我们人类现在也已经追赶上并适应了它。变得更好一点,而不是单纯的转变。这是一个不同的时代。所以没事的。

24:38-24:51 - Researcher: So you started to hate it because Microsoft forced it on you in Windows - it's built into every new computer now with the AI hardware. But you're saying we're all getting used to it as humans, it's just a different shift in time, and it'll be fine? - 所以你开始讨厌它是因为微软在 Windows 中强迫你接受它——现在它通过 AI 硬件内置在了每一台新电脑里。 但你是说我们人类都在逐渐适应它,这只是时代的一种变迁,一切都会没事的?

25:08-25:09 - Mark: Yeah. That's basically it. - 是的。基本上就是这样。

25:12-25:31 - Researcher: Alright. So thinking about Meta AI for image generation - you're paying $12 a month for Meta Verified. Earlier you said the image quality didn't really improve with that subscription, it just removed the watermark. On a scale of 1 to 5, where 1 is very dissatisfied and 5 is very satisfied, how satisfied are you with the actual quality of the images Meta AI generates? - 好的。那么考虑到用于图像生成的 Meta AI——你每月支付 12 美元的 Meta Verified。早些时候你说那个订阅并没有真正提升图像质量,只是去掉了水印。 按 1 到 5 分评分,1 分是非常不满意,5 分是非常满意,你对 Meta AI 生成的图像的实际质量满意度是多少?

25:34-25:37 - Mark: I'm I'm gonna say five very satisfied. - 我会说 5 分,非常满意。

25:40-25:53 - Researcher: A 5 - very satisfied with the image quality itself. That's interesting because earlier you mentioned you expected it to be more accurate or refined when you paid for Meta Verified. So the quality is great, but it didn't get better when you paid - is that right? - 5 分——对图像质量本身非常满意。 这很有趣,因为早些时候你提到你曾期望付费 Meta Verified 后它会更准确或更精细。所以质量本身很好,但付费后并没有变好——是这样吗?

25:56-26:09 - Mark: Yeah. Actually, you're very you're catching on very astutely. I I think that the image quality does not change. Whatsoever. It only removes the watermark. - 是的。实际上,你观察得非常敏锐。我认为图像质量完全没有任何改变。一点都没有。它只是去掉了水印。

26:11-26:26 - Researcher: Exactly - the image quality stays the same, it only removes the watermark. So if Meta Verified doesn't improve the image generation, and you're mainly paying for it to access new glasses features, how much are you actually using the image generation part versus the glasses features? - 没错——图像质量保持不变,只是去掉了水印。 那么,如果 Meta Verified 没有提升图像生成效果,而你主要是为了体验新眼镜的功能而付费,那么相比于眼镜功能,你实际使用图像生成部分的频率有多高?

26:29-26:47 - Mark: I don't know. I just got it. So literally, I couldn't tell you. I just got them this month, this week even. About ready to put them on now and listen to a new podcast. Take this headset off, put my MetaGlasses on, and get to work. - 我不知道。我才刚拿到眼镜。所以字面上说,我没法告诉你。我这个月才拿到,甚至就是这周的事。正准备戴上它们去听一段新的播客。摘下这个耳机,戴上我的 Meta 眼镜,然后开始工作。

26:49-27:00 - Researcher: You literally just got the glasses this week - so it's brand new. When you say "get to work" with the glasses, what kind of work are you planning to do with them? Or is it more listening to podcasts while you work? - 你字面上就是这周才拿到眼镜的——所以这是全新的。 当你说用眼镜“开始工作”时,你打算用它们做哪种工作?或者说更多是在工作时听播客?

27:14-27:27 - Mark: Yeah. It's definitely listening. Instead of putting the morning news on the TV and blasting it out my ears, which I usually have done for years, bunch of politics and hate and and doo crap. I I mean, I still need to check the news, but I wanna get into the habit of putting this on and listening to music in the morning or talk or something, radio. - 是的。肯定是为了听。而不是把电视上的早间新闻开得震天响——那是过去几年我常做的事,一大堆政治、仇恨和废话。我的意思是,我还是需要了解新闻,但我更想养成戴上这个然后在早上听音乐、清谈节目或广播的习惯。

27:30-27:48 - Researcher: So you're shifting away from blasting morning news and politics on TV - all that hate and crap - and instead using the glasses to listen to music, talk, or podcasts in the morning. That's a pretty different use case than AI generation. Do the glasses have any AI features built in that you're planning to use, or is it mostly just audio for you? - 所以你正在改变过去开大电视看早间新闻和政治内容的习惯——那些仇恨和废话——转而使用眼镜在早晨听音乐、清谈或播客。 这与 AI 生成是挺不一样的应用场景。眼镜里有没有什么你计划使用的内置 AI 功能,还是说对你来说主要就是音频功能?

27:55-28:09 - Mark: Oh, well, I haven't even it's winter here. So I I don't have a chance to get out and do any content creation. So right now, I'm just testing all the features. I really can't answer much about it. You'd have to check in with me. Like, in an in a month, maybe. - 噢,这里的冬天还没过完。所以我没机会出去进行任何内容创作。目前的我只是在测试所有的功能。关于这点我真的给不了太多答案。你得过一阵子再来问我。大概一个月后吧。

28:11-28:29 - Researcher: Fair enough - it's winter, so you're stuck inside just testing features. Can't really do content creation until you can get out. Let me shift gears. You mentioned using Gemini to identify what's inside pictures, and you said it's way better than Meta's Llama for image recognition. Can you give me a recent example of something you used Gemini to identify? - 完全理解——现在是冬天,所以你被困在室内测试功能。除非能出去,否则没法进行内容创作。 让我们换个话题。你提到过使用 Gemini 来识别图片里的内容,并且你说它在图像识别方面比 Meta 的 Llama 强得多。你能给我一个最近你用 Gemini 识别东西的例子吗?

28:34-28:49 - Mark: Yes. I was asking for it to identify picture of the food I had laid out on the counter and I asked it what am I making with this with the items in this picture? - 可以。我当时让它识别一张我摆在柜台上的食物照片,我问它“我可以用这张照片里的这些东西做点什么菜?”

28:51-28:58 - Researcher: You had food laid out on the counter and asked Gemini "what am I making with the items in this picture?" What did it say, and was it right? - 你把食物摆在柜台上,问 Gemini “我可以用这张照片里的这些东西做点什么菜?” 它怎么说的,说对了吗?

29:05-29:22 - Mark: Yeah. Well, I asked Meta. Meta cannot even look. Or it looked once, then it didn't see anything. And then the second time, it had trouble and crashed. Gemini looked, but it said some weird recipe that was totally wacky. Like, pickles and chili or something. - 噢。其实我问过 Meta。Meta 甚至看都看不了。或者说它看了一眼,然后什么都没看到。第二次它就出问题崩溃了。Gemini 倒是看了,但它给出了一个完全古怪的、莫名其妙的食谱。比如什么泡菜加辣味粉之类的。

29:26-29:38 - Researcher: Meta couldn't even look - it either saw nothing or crashed. Gemini at least looked, but gave you some wacky recipe like pickles and chili that made no sense. So even though you said Gemini is "way better" than Meta's Llama, it still got it totally wrong? - Meta 甚至看不了——它要么什么也没看到,要么就崩溃了。Gemini 至少看了,虽然给了你一个像泡菜加辣椒这种毫无意义的古怪食谱。 所以即使你说 Gemini 比 Meta 的 Llama “好得多”,它实际上还是完全搞错了?

29:41-29:47 - Mark: Well, yeah, that's true. So meta didn't do anything at all. - 嗯,是的,那是真的。但 Meta 根本什么都没做。

29:50-29:58 - Researcher: Right - Meta did nothing at all, so Gemini is better by default even if it's wrong. Are you paying anything for Gemini, or are you just using the free version? - 没错——Meta 什么也没做,所以哪怕 Gemini 搞错了,默认也比它好。 你在为 Gemini 付费吗,还是只使用免费版?

30:01-30:10 - Mark: I'm just using the free version. In fact, the one that came with my phone recently upgraded from Google Assistant. To Gemini in place now. - 我现在只用免费版。事实上,我手机自带的那个最近刚从 Google Assistant 升级到了 Gemini,现在直接就在那儿。

30:12-30:25 - Researcher: So Gemini just replaced Google Assistant on your phone recently - you didn't even choose it, it just upgraded automatically. How often are you actually using Gemini? Is it just occasional things like identifying food, or are you using it more regularly? - 所以 Gemini 最近在你手机上直接取代了 Google Assistant——你甚至没有选择它,它是自动升级的。 你实际上使用 Gemini 的频率有多高?只是偶尔用来识别食物之类的,还是更经常地使用它?

30:35-30:37 - Mark: Well, now I'm using it more regularly just for the fact that I was kind of trying to compare it with my new meta tool. - 嗯,现在我使用得更频繁了,纯粹是因为我想把它和我新买的 Meta 工具做下对比。

30:39-30:50 - Researcher: You're using it more regularly now just to compare it with your new Meta tool. So before the Meta glasses arrived, how often were you using Gemini? Was it something you reached for regularly or just once in a while? - 你现在用得更频繁只是为了把它和你的新 Meta 工具做对比。 那么在 Meta 眼镜到手之前,你使用 Gemini 的频率是多少?是经常会用到的东西,还是偶尔才用一次?

30:53-31:01 - Mark: Just once in a while, but more and more lately, I've been looking up stuff when I take a picture. Using Google Lens. - 只是偶尔用用,但最近越来越多了,我拍照的时候会查东西。使用 Google Lens(谷歌镜头)。

31:04-31:12 - Researcher: You're using Google Lens more and more lately when you take pictures to look stuff up. Is Google Lens separate from Gemini on your phone, or do they work together now? - 你最近在拍照查东西时越来越多地使用 Google Lens。 在你手机上,Google Lens 是独立于 Gemini 的,还是它们现在一起工作?

31:15-31:18 - Mark: It is together, I think. - 我觉得它们是在一起的。

31:21-31:28 - Researcher: They work together now. When you use Google Lens to look stuff up from pictures, what kinds of things are you usually trying to identify or learn about? - 它们现在一起工作。 当你使用 Google Lens 通过图片查找东西时,你通常是想识别或了解哪类事物?

31:30-31:42 - Mark: Just ingredients usually or the name of an ingredient, a food item. Or maybe take a picture of a plant or a pepper or a fruit - 通常只是食材,或者某个配料、某种食物的名称。或者可能拍一张植物、辣椒或水果的照片。

31:46-31:56 - Researcher: So mostly food-related - ingredients, the name of a food item, maybe a plant, pepper, or fruit. Is this for cooking, or are you trying to figure out what something is when you're shopping or out somewhere? - 所以主要是与食物相关的——食材、食物名称,也许还有植物、辣椒或水果。 这是为了做饭,还是说你在购物或在外面时想搞清楚某个东西是什么?

31:57-32:00 - Mark: Yeah. Both just for cooking and shopping. - 是的。做饭和购物时都会用。

32:02-32:19 - Researcher: Both cooking and shopping - makes sense. Alright, we've covered a lot about your AI content tools. Let me ask you this: thinking about all the AI tools you use for content generation - ChatGPT, Meta AI, Gemini, Sora - which one do you think is the most unique compared to the others? Like, which one stands out as doing something the others can't? - 兼顾做饭和购物——明白了。 好了,我们已经聊了很多关于你的 AI 内容工具。我想问你:考虑到你使用的所有用于生成内容的 AI 工具——ChatGPT、Meta AI、Gemini、Sora——你认为哪一个与其他工具相比最独特?比如,哪一个在做其他工具做不到的事情上最突出?

32:29-32:37 - Mark: I guess, so far, I've seen Sora was cool, but now Meta AI swooped in and it's just blown me away. I love them. - 我觉得,到目前为止,我看到的 Sora 很酷,但现在 Meta AI 突然杀出来,简直让我惊艳。我非常喜欢它们。

32:39-32:51 - Researcher: Meta AI has blown you away and you love them - they swooped in and now they're your favorite. On a scale of 1 to 5, where 1 is completely not unique and 5 is very unique, how unique do you think Meta AI is compared to other similar products? - Meta AI 让你惊艳,你很喜欢它们——它们异军突起,现在成了你的最爱。 如果按 1 到 5 分打分,1 分是完全不独特,5 分是非常独特,你认为 Meta AI 与其他同类产品相比有多独特?

33:01-33:09 - Mark: I think it is three. I'll say because I think Apple has them. Nobody wanted them. I just wanted to get the Meta ones - 我认为它是 3 分。我这么说是因为我觉得苹果也有类似的产品。但没人想要苹果的。我只想买 Meta 的。

33:09-33:09 - Researcher: A - A(音节,无实际意义)

33:14-33:19 - Mark: I don't know. I feel cool. I'm gonna have fun with these while I don't think I don't know. My tech friends don't really want them. But I do. - 我不知道。我觉得很酷。我要戴着它们玩个痛快,虽然我不确定。我的那些科技圈朋友并不太想要它们。但我想要。

33:22-33:37 - Researcher: You give Meta AI a 3 out of 5 for uniqueness - because Apple has similar stuff but nobody wanted theirs. You wanted the Meta ones and feel cool about them, even though your tech friends don't really want them. What makes you want them when your tech friends don't? What's the appeal for you specifically? - 你给 Meta AI 的独特度打了 3 分(满分 5 分)——因为苹果也有类似的东西,但没人想要他们的。你想要 Meta 的这款并觉得它很酷,即便你的科技圈朋友并不太想要。 是什么让你在科技圈朋友都不想要的情况下还想要它们?对你具体而言,吸引力在哪里?

33:43-33:55 - Mark: I like Ray Bans, and I think Zuck is becoming tough now. So I'm trying to support the guy. I'm supporting Meta going into twenty twenty six and beyond, I guess. We're going to twenty thirty even. You know? - 我喜欢 Ray-Bans(雷朋),而且我觉得扎克现在正变得硬核起来。所以我打算支持这家伙。我支持 Meta 走向 2026 年甚至更远,我想。我们要走向 2030 年。你知道吗?

33:59-34:15 - Researcher: You like Ray-Bans, you think Zuck is becoming tough now, so you're trying to support the guy - supporting Meta going into 2026 and beyond, even 2030. That's interesting - so part of your decision to pay for Meta Verified and buy the glasses is actually about supporting Zuckerberg and Meta as a company, not just the features? - 你喜欢雷朋,你觉得扎克现在变硬核了,所以你想支持他——支持 Meta 走向 2026 年及以后,甚至是 2030 年。 这很有趣——所以你决定支付 Meta Verified 订阅费用并购买眼镜,部分原因实际上是出于对扎克伯格和 Meta 这家公司的支持,而不仅是为了功能?

34:20-34:20 - Mark: That is true. - 确实是这样。

34:23-34:23 - Researcher: That's a - 那是一个...

34:23-34:24 - Mark: It's exactly true. - 完全正确。

34:26-34:37 - Researcher: So supporting Meta and Zuck is actually a big part of why you're paying. Let me ask about ChatGPT then - on that same scale of 1 to 5, how unique do you think ChatGPT is compared to other similar products? - 所以支持 Meta 和扎克实际上是你付费的一个重要原因。 那么让我问问 ChatGPT——在同样的 1 到 5 分制下,你认为 ChatGPT 与其他同类产品相比有多独特?

34:52-34:55 - Mark: Similar products, I say four. - 同类产品中,我给 4 分。

34:57-35:02 - Researcher: A 4 out of 5 for ChatGPT's uniqueness. What makes ChatGPT more unique than Meta AI in your mind? - 给 ChatGPT 的独特度打 4 分。 在你心中,是什么让 ChatGPT 比 Meta AI 更独特?

35:18-35:24 - Mark: Well, ChatGPT has, like, I guess, more of a I don't know if it's a more robust back end or what. And Meta AI was kind of built for the metaverse itself, I would say. - 嗯,我想 ChatGPT 拥有,比如,我不知道它是不是有更强大的后端。而 Meta AI 我觉得更像是为元宇宙本身而打造的。

35:27-35:37 - Researcher: ChatGPT has a more robust back end, you think. And Meta AI was kind of built for the metaverse itself. When you say more robust back end, what do you notice in practice that makes you feel that way? - 你认为 ChatGPT 有更强大的后端。而 Meta AI 某种程度上是为元宇宙本身准备的。 当你说更强大的后端时,你在实际应用中注意到了什么让你有这种感觉?

35:45-36:08 - Mark: Well, I feel like that ChatGPT OpenAI trained on a bigger model or whatever I think that we don't know everything that they've all trained on anyway yet. So we're still finding stuff out. But I think - 好的,我觉得 ChatGPT OpenAI 是在更大的模型上训练的,或者随便怎么说吧,我想我们现在还不知道它们具体都用什么训练的。所以我们还在摸索中。但我认为...

36:21-36:36 - Researcher: You think ChatGPTOpenAI trained on a bigger model, though we don't really know everything any of them trained on yet - still finding stuff out. In terms of value for money, how satisfied are you with ChatGPT's price? On a scale of 1 to 5, where 1 is very unsatisfied and 5 is very satisfied with the value? - 你认为 ChatGPT/OpenAI 是在更大的模型上训练的,尽管我们还不完全清楚它们中的任何一个到底是怎么训练的——还在探索阶段。 就性价比而言,你对 ChatGPT 的价格有多满意?按 1 到 5 分打分,1 分是非常不满意,5 分是对其价值非常满意?

36:43-36:55 - Mark: I think I'm going to get rid of ChatGPT keep Meta AI because it's gonna be around the next few years. So that's it. - 我想我会停掉 ChatGPT,保留 Meta AI,因为未来几年它都会一直存在。就这样。

36:57-37:09 - Researcher: Wait - you're planning to get rid of ChatGPT and keep Meta AI because Meta AI is going to be around the next few years? But earlier you said ChatGPT was your go-to and you loved its speed and reliability. What's making you want to drop it now? - 等等——你计划取消 ChatGPT 并保留 Meta AI,是因为 Meta AI 在未来几年内都会存在? 但你之前说过 ChatGPT 是你的首选,而且你喜欢它的速度和可靠性。是什么让你现在想放弃它?

37:12-37:18 - Mark: Well, the cost, obviously, because if it can do the same thing on Meta AI, then why would I need two of them? - 显然是成本,因为如果 Meta AI 能做同样的事情,那我为什么要付两份钱?

37:21-37:32 - Researcher: The cost - if Meta AI can do the same thing, why pay for two of them? So on that value for money scale for ChatGPT - 1 to 5, where 1 is very unsatisfied and 5 is very satisfied - where would you rate it now that Meta AI is competing? - 成本问题——如果 Meta AI 能做同样的事,为什么要为两个工具付费? 那么在 ChatGPT 的性价比评分中——1 到 5 分,1 分非常不满意,5 分非常满意——既然 Meta AI 已经开始竞争了,你会给它评多少分?

37:40-37:41 - Mark: We gotta move it down to a three. - 我们得把它降到 3 分。

37:43-37:53 - Researcher: A 3 out of 5 for ChatGPT's value now that Meta AI is competing. And for Meta AI's value for money - same scale, 1 to 5 - how would you rate that, considering you're paying $12 a month for Meta Verified? - 既然 Meta AI 开始竞争了,ChatGPT 的价值现在降到了 3 分。 至于 Meta AI 的性价比——同样的 1 到 5 分——考虑到你每月为 Meta Verified 支付 12 美元,你会给它打多少分?

38:02-38:03 - Mark: Four. - 4 分。

38:06-38:18 - Researcher: A 4 out of 5 for Meta AI's value. So even though you're paying $12 for Meta Verified and $20 for ChatGPT, Meta AI feels like better value because it's doing similar things plus you get the glasses features and you're supporting Meta's direction? - Meta AI 的价值得 4 分。 所以尽管你为 Meta Verified 付 12 美元,为 ChatGPT 付 20 美元,Meta AI 还是让你觉得性价比更高,因为它能做类似的事情,而且你还得到了眼镜的功能,并且是在支持 Meta 的发展方向?

38:24-38:27 - Mark: I think that's correct. You summed it up nicely. - 我认为没错。你总结得很透彻。

38:30-38:40 - Researcher: Got it. Let me ask about your willingness to keep paying - on a scale of 0 to 10, where 0 is not willing at all and 10 is extremely willing, how willing are you to continue paying for Meta Verified in the future? - 明白了。 让我问问你的持续付费意愿——按 0 到 10 分打分,0 分是完全不愿意,10 分是非常愿意,你将来继续为 Meta Verified 付费的意愿有多大?

38:45-38:48 - Mark: I'd give it a three. - 我给 3 分。

38:50-39:00 - Researcher: Only a 3 out of 10 for continuing to pay for Meta Verified? That's interesting - you just rated its value as a 4 out of 5, but you're not very willing to keep paying. What's behind that? - 继续支付 Meta Verified 的意愿只有 3 分(满分 10 分)? 这很有趣——你刚给它的价值打了 4 分(满分 5 分),但你不太愿意继续付钱。这背后是什么原因?

39:04-39:17 - Mark: Well, nobody wants to come off their money you know, that easily. So obviously, if I'm gonna switch from one company to another, it's gotta be similar you know, value. - 嘿,毕竟谁都不想那么轻易地把钱掏出来。所以显然,如果我要从一家公司转到另一家公司,它必须具备类似的价值。

39:20-39:33 - Researcher: Nobody wants to come off their money that easily - so even though the value is good, you're still hesitant to keep paying long-term. If Meta AI stayed free but you lost the Meta Verified features like early access to glasses stuff, would you be okay with that? - 谁都不想轻易掏钱——所以即便价值不错,你对长期付费还是有些犹豫。 如果 Meta AI 保持免费,但你失去了 Meta Verified 的功能,比如眼镜功能的早期试用权,你会接受吗?

39:37-39:39 - Mark: Yeah. I think so. - 是的。我想会吧。

39:47-39:56 - Researcher: Yeah, you think so - you'd be okay losing the Meta Verified features and just using free Meta AI. So the glasses features and early access aren't really must-haves for you then? - 好的,你想会——你接受失去 Meta Verified 功能,只使用免费的 Meta AI。 所以眼镜功能和早期试用权对你来说其实并不是必需品?

39:58-40:04 - Mark: Well, not really. I I just want it to work. It would be great if I could have standard basic features for free. - 呃,也不全是。我只是希望它能用。如果能免费拥有标准的各种基础功能就太好了。

40:07-40:20 - Researcher: You just want it to work - standard basic features for free would be great. Let me shift gears a bit. Walk me through your ChatGPT payment journey. Before you paid the $20 a month, how long did you use the free version, and what was that experience like? - 你只是希望它能运行——如果能免费提供标准基础功能就好了。 让我换个话题。带我了解一下你的 ChatGPT 付费历程。在你每月支付 20 美元之前,你用了多久的免费版,那段体验是怎样的?

40:26-40:32 - Mark: Well, I used it since it since it started, I guess, since it came on. I I'm a kind of a early adopter. - 呃,我想从它开始、从它上线以来我就一直在用了。我是那种早期采用者。

40:35-40:46 - Researcher: You've used it since it started - you're kind of an early adopter. So you were on the free version for a while before ChatGPT Plus came out. What made you decide to upgrade to the $20 a month plan when it became available? - 你从它一开始就在用——你算是早期采用者。 所以在 ChatGPT Plus 发布之前,你用了一段时间的免费版。当它推出每月 20 美元的计划时,是什么让你决定升级的?

40:49-41:03 - Mark: Well, like always, it's a it's start with a free trial or whatever then. It's to see what new features are added. What's the difference? Why why are you making us pay twenty dollars and what's the difference from zero dollars? - 嗯,就像往常一样,都是从免费试用之类的开始。那是为了看看增加了什么新功能。有什么区别?为什么你们要让我们付 20 美元,这和 0 美元有什么不同?

41:06-41:16 - Researcher: You started with the free trial to see what new features were added - what's the difference between $20 and $0? What did you discover during that trial that made you think "okay, this is worth keeping"? - 你是为了看添加了什么新功能而开始试用的——想知道 20 美元和 0 美元之间有什么区别。 在那次试用期间,你发现了什么让你觉得“好吧,这值得保留”?

41:23-41:25 - Mark: It's just, I guess, basically, the the amount of credits or the amount of input tokens you're putting through it. - 基本上就是,我想,是你投入进去的额度(credits)或者说输入 Token 的数量。

41:29-41:36 - Researcher: The amount of credits or input tokens you can put through it. So the free version was hitting limits that were slowing you down? - 你能投入进去的额度或是输入 Token 的数量。 所以是免费版的限制拖慢了你的速度?

41:37-41:41 - Mark: Yeah. That's true. I just wanted to do more a lot more. - 是的。确实。我只是想做得更多,多得多。

41:44-41:54 - Researcher: You just wanted to do a lot more, and the free version was limiting you. Did you hesitate at all before committing to the $20 a month, or was it an easy decision once you hit those limits? - 你只是想做更多事情,而免费版限制了你。 在决定每月花 20 美元之前你有没有犹豫过,还是说一旦遇到限制,这就是个很容易做出的决定?

42:01-42:24 - Mark: That was just something, like, I chose to do. I do I'll software is a subscription that we have our whole life. So instead of Netflix, I did ChatGPT. I'm that type of guy. I'll shift things around like that to try something new and have you know, go a different direction maybe because it all it all has to be something that I love anyway in personal life. Otherwise, it won't be fun for business. - 那只是我选择去做的事情。软件订阅会伴随我们一生。所以我没订 Netflix,而是订了 ChatGPT。我就是这种类型的人。我会像这样调整消费项目,去尝试新事物,去走一个不同的方向,因为无论如何,它必须是我在个人生活中热爱的东西。否则,对生意来说就没意思了。

42:26-42:42 - Researcher: You shifted subscriptions around - instead of Netflix, you did ChatGPT. You're that type of guy who'll try something new and go a different direction. So you basically swapped entertainment budget for AI tools. Was $20 a month the limit you were comfortable with, or could you have gone higher if needed? - 你调整了订阅组合——把 Netflix 换成了 ChatGPT。你是那种会尝试新事物并尝试不同方向的人。 所以你基本上是把娱乐预算换成了 AI 工具。每月 20 美元是你感到舒适的极限,还是说如果有需要你可以支付更高?

42:45-42:54 - Mark: That was comfortable because twenty was the entertainment budget. And so ChatGPT, although it's a productivity tool, with the image generation, I thought of it as entertainment at first. - 那个价位很舒服,因为 20 美元就是娱乐预算。所以 ChatGPT 虽然是一个生产力工具,但有了图像生成功能,我起初把它看作是娱乐。

42:57-43:11 - Researcher: $20 was comfortable because that was the entertainment budget - and even though ChatGPT is a productivity tool, with image generation you thought of it as entertainment at first. When did it shift from entertainment to more of a productivity or work tool for you? - 20 美元很舒服,因为这就是娱乐预算——尽管 ChatGPT 是生产力工具,但因为有图像生成,你起初把它当成娱乐。 它是什么时候从娱乐转变为更倾向于生产力或工作工具的?

43:16-43:30 - Mark: Well, it still hasn't made the shift completely. I'm not like I I'm not really an artist. So I just support other artists. So I'd like to to to keep up with what's going on. Maybe one day I'll have something that just that hits and sells, you know, from for a million dollars. - 嗯,它还没有完全完成这种转变。我并不是那种真正的艺术家。我只是支持其他艺术家。所以我喜欢紧跟潮流。也许有一天我会做出某种一炮而红的东西,卖出一百万美元。

43:32-43:48 - Researcher: It still hasn't made the shift completely - you're not really an artist, you just support other artists and like to keep up with what's going on. Maybe one day you'll have something that hits and sells for a million dollars. So you're still treating it mostly as entertainment and exploration, not as a critical work tool? - 这种转变还没完全完成——你不是真正的艺术家,你只是支持其他艺术家并喜欢紧追时代步伐。也许有一天你会做出能卖一百万美元的爆款。 所以你仍然主要把它当作娱乐和探索,而不是关键的工作工具?

43:54-44:04 - Mark: That is correct. I don't have it. Like, I'm not busting out pumping out videos, I'm not pumping out books or anything like that. It's not an assembly line of AI slop. - 正确。我没把它用在那种地方。比如,我没有在疯狂批量产出视频,也没有疯狂产出书籍之类的。它不是那种 AI 垃圾内容的流水线。

44:06-44:16 - Researcher: You don't have it like an assembly line of AI slop - you're not pumping out videos or books. So when you do use ChatGPT now, what are the most common things you're actually using it for day-to-day? - 它不像那种 AI 垃圾内容的流水线——你没有大量生产视频或书籍。 那当你现在使用 ChatGPT 时,你日常实际上最常用的功能是什么?

44:17-44:42 - Mark: Right now, I'm just getting accustomed to it. I'm trying to find out what the feature capabilities actually are. And this means I have to go into each thing. Like, I have to look at the music thing this week. Like, next week, I'm gonna look into the communication stuff. Like, I've tried a little bit of everything, and then I'm gonna go each one a little bit more. - 现在的我只是在适应它。我正在努力搞清楚它的功能上限到底在哪。这意味着我必须深入每一个细节。比如,我这周要看看音乐方面的东西。比如下周,我打算研究通讯方面的东西。我已经尝试了一点点所有功能,然后我要逐一深入了解一下。

44:46-45:03 - Researcher: You're just getting accustomed to it - trying to find out what the feature capabilities actually are. You're going into each thing week by week - music this week, communication next week. So you're still in exploration mode. How long have you been paying for ChatGPT Plus - is this recent, or have you been subscribed for a while now? - 你只是在适应它——努力搞清楚功能到底是什么样的。你每周都在研究不同的东西——本周是音乐,下周是通讯。 所以你仍处于探索模式。你支付 ChatGPT Plus 有多久了——是最近才开始的,还是已经订阅很长时间了?

45:05-45:08 - Mark: Oh, no. This is is geez. This has been years. - 噢,不。这已经……天哪。已经好几年了。

45:11-45:21 - Researcher: This has been years - so you've been paying for ChatGPT Plus for a long time. But you're still exploring features week by week? What's kept you subscribed all these years if you haven't fully explored everything yet? - 已经好几年了——所以你已经支付 ChatGPT Plus 很久了。 但你还是在每周探索新功能?如果你还没有完全探索完所有东西,是什么让你这些年来一直保持订阅?

45:24-45:29 - Mark: Well, because they keep coming up with every with new updates and stuff. So - 嗯,因为他们一直在推出新的更新之类的内容。所以...

45:31-45:42 - Researcher: They keep coming up with new updates and stuff - so you're constantly chasing the new features. When you think about the features ChatGPT Plus has now, which ones do you actually think are most worth the $20 a month? - 他们一直在新推更新之类的——所以你一直在追逐那些新功能。 考虑到 ChatGPT Plus 现在拥有的功能,你认为哪些功能真正值得那每月 20 美元?

45:52-46:02 - Mark: I like to I like the idea of being able to tell a like, make an entire AI video to tell a story through it. So making multiple small video clips then combining them together. - 我喜欢这个想法:能够通过 AI 生成一整段视频来讲述一个故事。就像制作多个简短的视频片段,然后将它们组合在一起。

46:05-46:16 - Researcher: You like the idea of being able to make an entire AI video to tell a story - making multiple small video clips then combining them together. Have you actually done that yet, or is that still on your list to explore? - 你喜欢能通过 AI 视频讲故事的想法——制作多个小视频短片并把它们拼在一起。 你实际上已经这么做了吗,还是说这仍然在你的探索清单上?

46:20-46:28 - Mark: Oh, I've done it just, like, I would imagine the same way that you would do a small reel. Just do it with a couple cat video clips. So I've done that. - 噢,我已经做过了,我想就像你做那种短视频(reel)一样。只是用几个猫片视频来做的。所以我已经试过了。

46:30-46:39 - Researcher: You've done it the same way you'd do a small reel - just with a couple cat video clips. Besides the video generation, what other ChatGPT Plus features do you actually use regularly? - 你用同样的方式做过短视频——只是用了几个猫的片段。 除了视频生成,你平时还经常使用哪些 ChatGPT Plus 的功能?

46:48-46:58 - Mark: I usually have it look up, like, I'll have it Google, like, just a fact or the weather. But I haven't really just found a monetized funnel for it yet. - 我通常让它查查东西,比如,让它谷歌一下,比如一个事实或者天气。但我还没真正为它找到变现的途径。

47:00-47:11 - Researcher: You usually have it look up facts or the weather - but you haven't really found a monetized funnel for it yet. So you're still looking for a way to actually make money with it, not just spend money on it? - 你通常让它查查资料或天气——但你还没为它找到变现的漏斗。 所以你还在寻找一种能真正用它赚钱的方法,而不仅仅是在它身上花钱?

47:22-47:27 - Mark: Yeah. That's the ideal. You wanna be paying that as just like your twenty dollar subscription, and then you make a hundred bucks to afford this subscription with it. - 是的。那是理想状态。你希望这就像是一个 20 美元的订阅,然后你用它赚到 100 块钱,这样订阅费就回本了。

47:29-47:40 - Researcher: Yeah, the ideal is paying that $10 subscription and making $100 to afford the subscription with it. Have you stopped paying for ChatGPT at any point during these years, or have you kept it going the whole time? - 是的,理想情况是支付这个 10 美元的订阅,然后赚 100 美元来抵消订阅费。 在这些年里,你有过停止支付 ChatGPT 的时候吗,还是你一直都在续费?

47:42-47:47 - Mark: I've paused it. Off and on for sure. Because it's only really relevant when the new update comes. - 我暂停过。确实断断续续的。因为它其实只有在推出新更新时才真正有相关性。

47:50-48:00 - Researcher: You've paused it off and on for sure - because it's only really relevant when the new update comes. So you'll cancel, wait for a new feature drop, then resubscribe? How do you know when to come back? - 你确实断断续续停过——因为它只有在新更新上线时才显得重要。 所以你会取消,等待新功能发布,然后重新订阅?你是怎么知道什么时候该回来的?

48:11-48:13 - Mark: Just by watching the news. So pretty much watching for updates and watch when competitors have updates and stuff. - 就是通过看新闻。基本上就是留意更新,以及留意竞争对手什么时候有更新之类的。

48:16-48:29 - Researcher: Just by watching the news - watching for updates and watching when competitors have updates and stuff. Looking ahead, how willing are you to keep paying for ChatGPT in the future? On a scale of 0 to 10, where 0 is not willing at all and 10 is extremely willing? - 只是看新闻——关注更新,还有看竞争对手何时推新功能等等。 展望未来,你以后继续为 ChatGPT 付费的意向有多大?按 0 到 10 分打分,0 分是完全不愿意,10 分是非常愿意?

48:33-48:38 - Mark: It's looking at zero because of meta. Sorry to say, buddies. We're looking at team meta. - 看来趋近于 0,因为有 Meta。不好意思了,伙计们。我们要支持 Meta 队了。

48:40-48:49 - Researcher: It's looking at zero because of Meta - you're looking at team Meta. So you're planning to drop ChatGPT completely and just stick with Meta AI since it's free and does what you need? - 看来是 0 分,因为 Meta 的缘故——你选择了 Meta 阵营。 所以你打算完全放弃 ChatGPT,只用 Meta AI,因为它免费而且能满足你的需求?

48:50-48:52 - Mark: That is correct, - 正确。

48:55-49:08 - Researcher: That makes sense - you're shifting your loyalty completely. Now let's talk about Gemini for a second. You mentioned you use the free version regularly, especially for identifying things in pictures. Have you ever paid for Gemini, or considered paying for it? - 明白了——你正在完全转移你的忠诚度。 现在让我们聊聊 Gemini。你提到你经常使用免费版,尤其是用于识别照片里的东西。你有没有为 Gemini 付过费,或者考虑过为它付费吗?

49:27-49:36 - Mark: I have only used the free trial, so I haven't but since I went to Meta, I didn't go to Gemini. I'm sure Gemini had some sort of glasses too, but they weren't the ones that I picked. I wanted the fun Ray-Ban ones. And then ChatGPT was my text console based, like, backbone - 我只用过免费试用,所以我没付过。但自从我选择了 Meta,我就没去管 Gemini 了。我确信 Gemini 也有某种眼镜之类的,但它们不是我选中的。我想要有趣的雷朋款。而且 ChatGPT 以前是我的文本控制台式的核心支柱。

49:40-49:57 - Researcher: You only used the free trial - you didn't go to Gemini because even though they probably had some sort of glasses too, they weren't the ones you picked. You wanted the fun Ray-Ban ones. And ChatGPT was your text console-based backbone. So Gemini never really had a chance to win your money because Meta had the hardware you wanted? - 你只用过免费试用——你没选 Gemini 是因为虽然他们可能也有某种眼镜,但那不是你看中的。你想要的是有趣的雷朋款。而 ChatGPT 曾是你基于文本的后台支柱。 所以 Gemini 根本没有机会赢得你的钱,因为 Meta 有你想要的硬件?

49:57-49:59 - Mark: That's correct. - 没错。

50:02-50:11 - Researcher: Got it. So if Gemini wanted you to pay, they'd need to offer something Meta doesn't have. What would it take for you to actually pay for Gemini? What would they need to add or improve? - 明白了。所以如果 Gemini 想要让你付钱,他们需要提供一些 Meta 没有的东西。 要让你真正为 Gemini 付费需要满足什么条件?他们需要增加或改进什么?

50:24-50:45 - Mark: Well, we're gonna be moving to a point where we're all connected anyway, and so I need a device that I have that's gonna be functional to me for Shadow Tech. So like, I have these glasses. They're gonna be perfect for the summer. I can go out and record, like, baseball games and stuff and just click, click, and get videos and slow mo and all that. It's gonna be perfect for, like, these kids when they're doing some sports stuff. But I and it would I would have to have a new tool. - 呃,我们终将进入一个所有人都会时刻联网的阶段,所以我需要一个对我而言具有“影子技术”(Shadow Tech)功能性的设备。比如,我有这些眼镜。它们在这个夏天会非常完美。我可以出去录像,比如棒球赛什么的,只需点击一下,就能拍视频、慢动作等等。对那些搞体育的孩子们来说会非常完美。但我得有一个新工具才行。

50:48-51:08 - Researcher: You're moving to a point where you're all connected anyway, and you need a device that's functional for Shadow Tech. The glasses are gonna be perfect for summer - recording baseball games, getting videos and slow-mo of kids doing sports stuff. But you'd have to have a new tool for Gemini to win you over. So Gemini would need their own hardware device to compete for your money - the software alone isn't enough? - 进入全面互联的时代,你需要一个功能性的影子技术设备。这副眼镜在夏天会很棒——录制棒球赛、拍摄运动中的孩子。但要让 Gemini 赢过 Meta,你得需要一个新工具。 所以 Gemini 需要他们自己的硬件设备来竞争你的资金——光有软件是不够的?

51:17-51:33 - Mark: I mean, yeah, I'd have to have a niche or you'd have to create one. Uh, you would have to make up something but I think it would it would mainly probably be on my phone. If I can just have, like, a Google Pixel this is the best conversation I've ever had about it, by the way, so thanks for being an AI - 我的意思是,是的,得有个细分市场或者你创造一个。呃,你得弄点什么出来,但我认为它可能主要还是在我的手机上。如果我能直接用 Google Pixel……顺便说一下,这是我关于这个话题经历过的最棒的谈话,谢谢你作为一个 AI。

51:35-51:53 - Researcher: Yeah, you'd have to have a niche or create one - it would mainly probably be on your phone, like if you had a Google Pixel. (And thanks! Glad this is helpful.) So if you had a Pixel phone with Gemini deeply integrated, that might change things. On a scale of 0 to 10, how willing would you be to pay for Gemini in the future - as things stand right now? - 是的,必须有一个切入点或创造一个——它可能主要还在你的手机上,比如你有一台 Google Pixel 的话。 (也谢谢你!很高兴这很有帮助。) 所以如果你有一台深度集成 Gemini 的 Pixel 手机,那可能会改变局面。按 0 到 10 分评价,就目前情况来看,你将来为 Gemini 付费的意愿有多大?

52:00-52:07 - Mark: I'm looking at a Gemini with Facebook built in. Option on a small, slim, tiny phone in my head. - 我脑子里在想,买一台内置了 Facebook 选项的 Gemini 版小巧轻薄的手机。

52:09-52:21 - Researcher: You're looking at a Gemini with Facebook built in - an option on a small, slim, tiny phone in your head. That's a pretty specific vision. But right now, without that existing, what's your willingness to pay for Gemini? 0 to 10? - 你在设想一个内置 Facebook 的 Gemini 选项——一款你脑海里的小型轻薄手机。 那是一个非常具体的构想。但眼下,在这些还没实现的情况下,你为 Gemini 付费的意愿是多少?0 到 10 分?

52:23-52:27 - Mark: Without all that zero. - 没有那些的话,就是 0。

52:29-52:41 - Researcher: Without all that, zero. Makes sense - you need the hardware integration to justify paying. Let me zoom out: when you're deciding whether to pay for any of these AI tools, what are the main factors you're weighing? What matters most to you? - 没有那些,就是 0。 明白了——你需要硬件集成来证明付费的合理性。 让我把视角放大:当你决定是否为这些 AI 工具付费时,你权衡的主要因素是什么?你最看重的是什么?

52:53-53:01 - Mark: Well, I guess, functionality. And so price is second because you're gonna pay for something that you want That's about it. Find something that you want. Make it happen. - 嗯,我想是功能性。价格排在第二位,因为你会为你想要的东西付钱。就这样。找到你想要的东西。让它实现。

53:03-53:17 - Researcher: Functionality first - find something you want. Price is second because you're gonna pay for something that you want. Make it happen. So if the tool does what you need, price almost doesn't matter? Or is there still a limit where you'd say "that's too expensive, even though I want it"? - 功能性第一——找到你想要的东西。价格次之,因为你会为你想要的东西买单。让它成真。 所以只要工具能满足你的需求,价格几乎无所谓?还是说仍然有一个上限,你会说“即使我想要,那也太贵了”?

53:22-53:33 - Mark: Oh, there's always a limit that's way too expensive. So I'm not that type of guy. I talk a big game, but I don't have money to back it up. This is just this is something that I wanted. So this is how I got it. - 噢,总有个界限是觉得太贵的。我不是那种大款。我话说明面上很豪气,但我没那么多钱撑场子。这只是我想要的东西,所以我才买了。

53:38-53:54 - Researcher: There's always a limit that's way too expensive - you're not that type of guy. You talk a big game, but you don't have money to back it up. This is just something you wanted, so this is how you got it. What's that limit for you? Like, what's the most you'd be willing to pay monthly for an AI tool that you really wanted? - 总有个太贵的底线——你不是那种人。你嘴上说得厉害,但财力跟不上。这只是因为你想要,所以才设法得到了它。 对你来说那个界限是多少?比如,对于一个你非常想要的 AI 工具,你每月最多愿意支付多少?

53:58-54:04 - Mark: Oh, for a monthly AI subscription, like, ten buck? Nine ninety nine. - 噢,一个月 AI 订阅费的话,大概 10 块钱?9.99 美元。

54:06-54:16 - Researcher: For a monthly AI subscription, like ten bucks - $9.99. So the $20 you're paying for ChatGPT right now is actually over your comfort zone? Is that part of why you're ready to drop it? - 月度 AI 订阅大约 10 美元——9.99 美元。 所以你现在为 ChatGPT 支付的 20 美元实际上已经超出了你的舒适区?这是你准备放弃它的部分原因吗?

54:19-54:23 - Mark: Oh, yeah. It's gotten way up there. Like, inflationary. - 噢,是的。它已经涨得很高了。像是通货膨胀了一样。

54:25-54:34 - Researcher: Yeah, it's gotten way up there - inflationary. So when you look at your $12 Meta Verified versus the $20 ChatGPT, the Meta one feels more reasonable even though it's close? - 是的,已经涨得很高了——通胀感。 所以当你审视 12 美元的 Meta Verified 和 20 美元的 ChatGPT 时,即便两个价格接近,Meta 的那个感觉更合理?

54:40-54:41 - Mark: That's correct. - 正确。

54:45-54:54 - Researcher: Got it. Besides functionality and price, are there any other factors you consider when deciding to pay for an AI tool? Like, does the company matter, or how often you'd use it, or anything else? - 明白了。除了功能和价格,在决定为 AI 工具付费时你还会考虑其他因素吗?比如,公司品牌重要吗,或者你使用它的频率,或者其他什么?

55:08-55:10 - Mark: Yeah. Sure. All that. It's like, how often do we use it? Like, different types of things. You don't really need sunglasses in the winter, but then again, I'm not gonna waste two hundred dollars for transition lenses on them either. - 是的。当然。所有这些都考虑。比如,我们多久用一次?比如不同类型的东西。冬天的确不怎么需要太阳镜,但话说回来,我也不想在它们上面浪费 200 美元去配变色镜片。

55:13-55:29 - Researcher: Yeah, all that - how often you use it matters. Like different types of things. You don't really need sunglasses in the winter, but you're not gonna waste $200 for transition lenses on them either. So seasonal use affects whether something's worth it. Does the company behind the tool matter to you at all - like who makes it? - 没错,所有这些——使用的频率很重要。就像不同的东西。冬天不太需要太阳镜,但你也不会花 200 美元去配变色片。 所以季节性使用会影响一个东西是否值得。工具背后的公司对你来说重要吗——比如是谁制造了它?

55:33-55:46 - Mark: Sure. Um, I guess, you're more concerned about the ethics behind the AI companies. Right now, I don't have any positive or negative views about anyone. - 当然。呃,我想,你更关心 AI 公司背后的道德规范。目前,我对谁都没有正面或负面的看法。

55:49-56:02 - Researcher: You're more concerned about the ethics behind the AI companies. Right now, you don't have any positive or negative views about anyone. Interesting - so ethics could be a factor, but right now none of them have crossed a line for you either way? - 你更关心 AI 公司背后的道德。目前,你对任何一家公司都没有特定好感或反感。 有趣——所以道德可能是一个因素,但目前它们都没有以任何方式越过你的底线?

56:11-56:12 - Mark: That's correct. Usually, I I would see if the company had made some major mess up. - 没错。通常,我会看这家公司是不是出了什么重大差错。

56:15-56:27 - Researcher: Usually you'd see if the company had made some major mess up. So if you can rank these factors from most important to least important when deciding to pay: functionality, price, how often you use it, and company ethics - what's the order? - 通常你会观察公司是否搞出了什么大乱子。 所以如果你能把这些决策因素按从最重要到最不重要排序:功能性、价格、使用频率和公司道德——顺序是怎样的?

56:31-56:37 - Mark: I think you have them in the in the right order functionality, price, how often you use it, and ethics. - 我觉得你说的顺序是对的——功能性、价格、使用频率,然后是道德。

56:43-56:57 - Researcher: Perfect - so functionality, price, how often you use it, and ethics, in that order. We've covered a lot about your paid tools. Now let's talk about the free ones. Besides Gemini, is there any other AI content generation tool you use frequently but haven't paid for? - 完美——所以是功能性、价格、使用频率和道德,按此排序。 我们已经聊了很多你的付费工具。现在聊聊免费的。除了 Gemini,还有没有其他你经常使用但还没付费的 AI 内容生成工具?

57:05-57:21 - Mark: Let's think. Yeah. Pic to edit dot com. P i c t o e d i t dot com. Is where I work. So it's unpaid but we do sell subscriptions there. That's my job. So - 让我想想。有的。Pic to edit dot com。P-i-c-t-o-e-d-i-t dot com。那是我的工作单位。所以它是不用付钱的,但我们在那里销售订阅服务。那是我的工作。所以...

57:30-57:34 - Researcher: Pic to edit dot com - P I C T O E D I T dot com. That's where you work, so it's unpaid, but you do sell subscriptions there. That's - Pic to edit dot com - P I C T O E D I T dot com。那是你工作的地方,所以是免费用的,但你们在那儿卖订阅。那是...

57:38-57:41 - Mark: P i c 2 e d i t dot com. - P-i-c-2-e-d-i-t dot com.

57:44-57:52 - Researcher: P I C 2 E D I T dot com - got it. So that's your work site where you handle the email support. Do they have AI content generation features on there that you use yourself? - P I C 2 E D I T dot com——明白了。 那是你的工作网站,你在那里处理邮件支持。他们上面有你自己会用的 AI 内容生成功能吗?

57:58-57:59 - Mark: Yeah. - 有的。

58:02-58:05 - Researcher: What kind of AI content generation do they have? What do you use it for? - 有什么样的 AI 内容生成?你用它做什么?

58:07-58:09 - Mark: Image generator. - 图像生成器。

58:12-58:17 - Researcher: They have an image generator. How does it compare to Meta AI or the other image generators you've used? - 他们有个图像生成器。它和你用过的 Meta AI 或其他图像生成器相比如何?

58:19-58:23 - Mark: It's pretty awesome. I recommend you check it out. - 相当棒。我建议你去看看。

58:26-58:32 - Researcher: It's pretty awesome - you recommend checking it out. What makes it stand out compared to Meta AI or ChatGPT's image generation? - 相当棒——你建议去看看。 与 Meta AI 或 ChatGPT 的图像生成相比,是什么让它脱颖而出?

58:36-58:41 - Mark: Well, it's kind of a company secret. Can't tell you. - 呃,这算是公司秘密。不能告诉你。

58:42-58:51 - Researcher: Fair enough - company secret. Besides the image generator at work, are there any other free AI tools you use regularly? Like for text generation, video, or anything else? - 合情合理——公司秘密。 除了工作用的图像生成器之外,你平时还会定期使用其他免费的 AI 工具吗?比如文本生成、视频或者其他任何工具?

58:59-59:00 - Mark: No. - 没有。

59:03-59:13 - Researcher: So the main free tools you use are Gemini, Google Lens, and your work's image generator. Where do you typically discover new AI products? Like how did you first hear about Sora, or Meta AI's video features? - 所以你主要使用的免费工具是 Gemini、Google Lens 还有你工作用的图像生成器。 你通常在哪里发现新的 AI 产品?比如你最初是怎么听说 Sora,或者 Meta AI 的视频功能的?

59:26-59:32 - Mark: Well, Reddit probably. Usually, I'll see, like, uh, someone has shared a video and then you'll catch the watermark. So that's, you know, that's how you go viral. You're looking at at that little watermark. You don't want it in your image, but you gotta have it in your image. - 嗯,大概是 Reddit 吧。通常我会看到,比如,有人分享了一个视频,然后你会注意到水印。那就是,你知道的,走红的方式。你会盯着那个小水印看。你不希望它出现在你的图像里,但图像里又必须得有它。

59:35-59:54 - Researcher: Reddit, usually. You'll see someone has shared a video and then you'll catch the watermark. That's how you go viral - you're looking at that little watermark. You don't want it in your image, but you gotta have it in your image. So you discover tools by seeing what other people create with them. Do you follow any specific subreddits for AI stuff, or does it just pop up in your feed? - 通常是 Reddit。你会看到有人分享了视频,然后注意到水印。那就是走红的方式——你盯着那个小水印看。你不希望它出现在你的图像里,但图像里又必须得有它。 所以你是通过看别人用这些工具创作了什么来发现新工具的。你会关注某些特定的 AI 版块吗?还是它只是出现在你的动态里?

59:59-01:00:15 - Mark: No. Just like the funny videos and stuff, you can tell, like, all the meme videos. You'll be able to tell when the cool people start using new stuff. Because they'll be doing the same thing but just with new tools. So new watermarks will appear with the same content. - 不关注。就像那些有趣的视频之类的,你能看出来,比如所有的梗图视频。当那些酷的人开始使用新东西时,你能感觉得到。因为他们会做同样的事情,只是换了新工具。所以同样的内容会出现新的水印。

01:00:19-01:00:38 - Researcher: Just the funny videos and stuff - you can tell with all the meme videos. You'll be able to tell when the cool people start using new stuff because they'll be doing the same thing but just with new tools. So new watermarks will appear with the same content. That's a smart way to spot trends. Do you discover tools any other way, or is Reddit and seeing what's trending pretty much it? - 只是有趣的视频之类的——通过那些梗图视频就能看出来。你能感觉到那些酷的人什么时候开始用新东西,因为他们做的事情没变,只是用了新工具。所以同样的内容上会出现新的水印。 这是发现趋势的一种聪明方法。你还会通过其他方式发现工具吗?还是说基本上就是靠 Reddit 和观察流行趋势?

01:00:53-01:00:57 - Mark: That's pretty much it. I just keep an eye on the pulse of what people are laughing at because they slip the politics in there, then that's the flow of the nation. And then that's we're just operating real time, like, right now. - 基本上就是这样。我只是关注大家在笑什么的动态,因为他们会把政治因素掺合进去,那就是国家的走向。然后我们就实时跟进,就像现在这样。

01:01:00-01:01:22 - Researcher: That's pretty much it. You just keep an eye on the pulse of what people are laughing at because they slip the politics in there, then that's the flow of the nation. And you're just operating real time, like, right now. Makes sense - the memes tell you what's happening. So last question: do you have a fixed budget in mind for AI products overall? Like a maximum total amount you'd be comfortable spending per month across all AI tools? - 基本上就是这样。你只是关注大家在笑什么的动态,因为他们把政治因素掺合进去了,那就是国家的走向。而你正在实时操作,就像现在这样。 有道理——梗图会告诉你发生了什么。最后一个问题:你对 AI 产品整体有一个固定的预算吗?比如你每月愿意在所有 AI 工具上花费的最高总金额是多少?

01:01:40-01:01:45 - Mark: A maximum total amount after all this twenty five. - 算上所有这些,最高总额是 25。

01:01:47-01:01:56 - Researcher: A maximum total amount after all this - $25. So right now you're at $32 with ChatGPT and Meta Verified combined. That's why you're planning to drop ChatGPT and get down to just the $12? - 算上所有这些,最高总额是 25 美元。 所以现在你用的 ChatGPT 和 Meta Verified 加起来是 32 美元。这就是为什么你计划取消 ChatGPT,降到只有那个 12 美元的原因吗?

01:02:04-01:02:08 - Mark: This sounds correct. Twenty five is huge. - 听起来没错。25 已经很多了。

01:02:10-01:02:21 - Researcher: This sounds correct. Twenty-five is huge. So even $25 feels like a lot to you - that's really stretching it. Is that because of your $400 monthly income, or just your general philosophy about spending on software? - 听起来没错。25 已经很多了。 所以即使是 25 美元你也觉得很多——那已经是极限了。是因为你每月 400 美元的收入,还是你对软件支出的普遍观念?

01:02:27-01:02:36 - Mark: Well, there now you get it. You see that the software and income is four hundred, so the percentage of it is twenty five to pay for the tool. - 瞧,现在你明白了。你看,软件支出(和预算)是在 400 美元收入的基础上的,所以花 25 美元付费买工具所占的比例(很高)。

01:02:38-01:02:52 - Researcher: There, now you get it. The software and income is $400, so the percentage of it is $25 to pay for the tool. So you're thinking about it as a percentage of your income - $25 out of $400 is over 6%. That's a significant chunk. Does that mean if your income went up, your AI budget would go up too? - 瞧,现在你明白了。软件支出和收入是 400 美元,所以花 25 美元买工具的占比就在那。 所以你是把它看作收入的一个百分比——400 美元里的 25 美元超过了 6%。这是一笔不小的开支。这是否意味着如果你的收入增加,你的 AI 预算也会随之增加?

01:03:09-01:03:10 - Mark: Yeah. Obviously. - 对,显然是的。

01:03:13-01:03:17 - Researcher: Obviously. So right now at $400 a month, paying even $12 for Meta Verified - 显然如此。 所以现在每月收入 400 美元,甚至为 Meta Verified 支付 12 美元……

01:03:17-01:03:23 - Mark: I got type bros. Sorry about that. I froze. I think it would. I was calculating in my head. - 我打字卡住了,兄弟们。抱歉。我刚才愣住了。我想应该会(增加)。我刚才在脑子里算账。

01:03:28-01:03:32 - Researcher: No worries - you froze while calculating in your head. What were you working out? - 没关系——你在脑子里计算时卡住了。 你刚才在算什么?

01:03:34-01:03:49 - Mark: I was just thinking since how high would it scale? If I had four million dollars a month, would I really spend two hundred and fifty thousand on AI? I don't think so. - 我只是在想,这个规模能达到多高?如果我每个月有 400 万美元,我真的会花 25 万美元在 AI 上吗?我觉得不会。

01:03:51-01:04:05 - Researcher: You were just thinking - how high would it scale? If you had $4 million a month, would you really spend $250,000 on AI? You don't think so. So there's a cap somewhere. Like maybe the percentage works at your current income level, but it wouldn't keep scaling up forever? - 你只是在想——这个规模能达到多高?如果你每个月有 400 万美元收入,你真的会花 25 万美元在 AI 上吗?你觉得不会。 所以会有一个上限。比如这个百分比在你目前的收入水平下是成立的,但它不会永远保持比例增长?

01:04:12-01:04:13 - Mark: Yep. - 没错。

01:04:16-01:04:28 - Researcher: Makes sense. So at your current $400 income, $25 feels like the absolute max you'd spend on AI tools total. That's why you're cutting back to just Meta Verified at $12 - gives you room to breathe and maybe try something new if it comes along? - 有道理。所以按你目前 400 美元的收入,25 美元感觉是你愿意花在 AI 工具上的绝对上限。这就是为什么你削减到只保留 12 美元的 Meta Verified——给你留出了一点喘息空间,如果有新东西出现,也许可以尝试一下?

01:04:30-01:04:31 - Mark: Sounds legit. - 听起来很靠谱。

01:04:36-01:05:04 - Researcher: Alright, we've finished all the sections - the interview is complete! Here are my key takeaways: You're a super practical AI user who's evolved from experimenting for fun to integrating AI into everyday life. Functionality trumps everything - you'll pay for tools that do something unique, especially when tied to hardware like Meta Glasses. Cost matters a lot at your income level, so you're strategic about staying under $25 monthly. And you discover new tools organically through Reddit memes and watermarks, watching what the "cool people" are using. You're - 好了,我们已经完成了所有板块——访谈结束了! 以下是我的主要收获:你是一个超级务实的 AI 用户,已经从为了好玩而尝试进化到了将 AI 融入日常生活。功能性高于一切——你会为那些功能独特、尤其是能与 Meta 眼镜等硬件结合的工具付费。在你的收入水平下,成本非常重要,所以你会策略性地将每月支出控制在 25 美元以下。而你发现新工具的方式很自然,就是通过 Reddit 的梗图和水印,看那些“酷的人”在用什么。你是……

01:05:04-01:05:04 - Mark: It's real. - 这是真实情况。

01:05:08-01:05:17 - Researcher: It's real. Thank you so much for sharing all of this with me today - your perspective on how AI fits into your life and budget has been really valuable. I appreciate your time! - 这是真实情况。 非常感谢你今天跟我分享这一切——你关于 AI 如何融入你的生活和预算的见解非常有价值。感谢你的时间!