Skip to main content

Command Palette

Search for a command to run...

I'm the Laziest Organizer on the Internet: While I Sleep, My Computer Labels 1000 Photos for Me

Updated
3 min read
I'm the Laziest Organizer on the Internet: While I Sleep, My Computer Labels 1000 Photos for Me

You know what I used to dread every morning?

Not waking up. Finding things.

Last week I wanted to find a photo from a coffee shop in Chiang Mai last year. I dug through folders, searched file names with Everything, went through my camera roll -- 45 minutes of hunting, and I couldn't even remember what the photo looked like.

I gave up.

Then I realized something: I haven't manually organized a single photo in three months.

My solution? Let my computer do the work while I sleep.

Here's how it works:

Step 1: Turn "I hate categorizing" into "I don't have to categorize"

I used to struggle with this. Every month I'd tell myself "this month I'll finally organize my photos." Every month I'd push it to next month, then give up entirely.

Now I don't categorize. I do one thing instead -- tag photos so AI understands what's in them.

My tool automatically analyzes every image on my computer. CPU usage? Basically zero. The moment I'm working, watching videos, or idle for more than 5 minutes, the background task kicks in.

What runs? OCR text recognition, face detection, scene classification.

A file named "IMG_20240115_083452.jpg" becomes "Chiang Mai coffee shop, brown cup, outdoor sunlight, wall graffiti." No action needed from me.

Step 2: Natural language search, replacing folder thinking

This is the part that surprised me most.

Before, to find "that photo from the Chiang Mai coffee shop," I had to remember which month, what the file name might contain, what folder I saved it in.

Now I just type in the search box: "brown cup sunlight graffiti wall outdoor cafe"

3 seconds. There it is.

The search logic changed -- not filename matching, but content understanding.

You don't need to remember file names. You just describe the scene you remember.

Step 3: Set it up once, then never think about it again

Configuring this system took me 30 minutes.

Now every morning I wake up with all new photos already processed. Tags applied, cloud synced, and I just open search.

Last week I traveled for three days and came back to 800 new photos. I let it run overnight while I slept. Next morning I opened the search, typed "airport lounge lights crowd," and found the photo I needed in 3 seconds.

My current organization logic:

  • Before: Organize photos -> sort into folders -> remember where you put them
  • Now: Take photos -> computer auto-tags -> search with natural language

You don't need to spend 30 minutes organizing every day. You only need to spend 30 seconds searching.

That's the real meaning of being "lazy" -- not doing nothing, but finding the right way to let machines handle all the repetitive work for you.

Time I used to spend "finding things": 45 minutes per day on average. Time I spend now: under 3 minutes.

The rest? Coffee.


rayslifelab.com

More from this blog

I Consumed AI Content for 2 Years. Then I Decided to Build Something With It

中文图文 我消费了两年 AI 工具后,决定用 AI 做点东西——然后做出来了 我大概在 2024 年初就开始用 ChatGPT 了。 那时候每天刷各种 AI 新闻、收藏教程、订阅 Newsletter,朋友圈分享 AI 工具的文章必点赞。我觉得自己很"懂" AI。 但有一天我仔细算了一下:两年来,我到底用 AI 做成了什么事? 答案是:0 件。 我让 AI 给我写过诗、润色过邮件、翻译过文档——这些有用,但它们本质上只是"用 AI 说话"。不是用 AI 做事。 我的转折点:那个我想解决的痛点...

Apr 28, 20261 min read
I Consumed AI Content for 2 Years. Then I Decided to Build Something With It

我每天生成30G的AI电影级视频-——我是如何防止自己被素材沉涂的?

我每天用 AI 生成30 G 的电影级视频——然后差点被自己的素材库涉汪。 事情是这样的。 大概两个月前,我开始用 Runway 和 Kling 生成 AI 电影短片。一开始很爱,但问题很快来了:每天产出几十 G 的视频碎片,一周下来硬盘就报警了。更要命的是,我想找上周做的那条"下雨的东京街头"——翻了三遍文件夹,找了两个小时,最后发现它被我误删进了"已处理"文件夹。 痛定思痛,我决定动手。 我的方案是这样的: 第一步:给所有视频自动打标签。 我写了个脚本,用 FFmpeg 截取视频的中间帧,然...

Apr 28, 20261 min read
R

Ray's blog

14 posts