Do You Think Everyone Needs to Grow Just Because You Do?

00:00 I was randomly added to a discussion group today.

00:05 The group's theme

00:07 was "Tech for Good."

00:10 I don't know why,

00:11 but the moment I saw the group,

00:13 I thought of "tech-enabled poverty relief."

00:17 Perhaps it reminded me

00:20 of how I felt back home over Lunar New Year.

00:24 At the time,

00:27 I met up with some former colleagues.

00:32 I realized their use of AI—

00:37 or their experience—stopped at Doubao and Qwen,

00:42 at a fairly superficial level.

00:46 They might use

00:49 these AI tools to ask questions.

00:51 But as for AI coding,

00:54 OpenClaw,

00:57 Claude Code, Cursor, and Gemini,

01:01 as well as ChatGPT,

01:04 they had barely used any of them.

01:08 Of course, I know

01:09 these tools have barriers to entry.

01:13 But my former colleague teaches IT

01:18 at a local private school.

01:23 And I wondered:

01:24 if even an IT teacher

01:26 hasn't encountered these tools,

01:29 what about their students?

01:33 I know many students are definitely ahead—

01:38 they move faster than their teachers

01:40 and encounter things

01:43 that are more in step with the times.

01:47 I just don't know whether those students

01:52 would discuss them with their teachers.

01:55 Or,

01:57 suppose the teacher

01:59 kept growing alongside

02:02 the changing times.

02:04 Wouldn't their students then

02:07 have far more room to grow?

02:10 I don't know.

02:12 So when I saw the group's name,

02:16 the idea of using technology

02:18 to alleviate poverty came to mind.

02:21 Sometimes I think,

02:25 rather than thinking about values around AI,

02:29 perhaps helping more people use AI

02:37 is more important.

02:40 But then I wonder:

02:43 if AI ethics and values

02:48 have no baseline,

02:51 no red lines,

02:53 and the rules for "Tech for Good"

02:58 aren't properly designed,

03:01 would helping people access AI

03:04 bring more benefits

03:06 or more harm?

03:11 I don't know.

03:14 So I talked to AI about it:

03:16 what exactly is "Tech for Good"?

03:22 It brought up several dimensions.

03:25 For example, when building an AI product,

03:29 what values guide it?

03:32 What personal data does it handle,

03:37 and how does it use that data?

03:40 How does the product guide its users?

03:45 These questions concern

03:50 the entrepreneur's side.

03:53 Beyond the CEO's values,

03:56 there are the product's values,

03:58 how the product uses

04:00 all the data users leave on the platform,

04:04 and what kind of impact

04:07 the product has on its users.

04:10 That led me to two other questions.

04:13 First, if a product truly aims to do good,

04:20 can it remain true from start to finish?

04:23 Before a product scales,

04:29 it may follow an ideal path.

04:31 But after it scales,

04:34 it may stray from its original direction.

04:37 That's one side of the issue.

04:41 It concerns

04:45 how the product aligns with its founder,

04:49 and whether that alignment

04:53 can survive future growth.

04:57 The second issue is capital allocation,

05:03 especially investors.

05:05 I just asked whether, after scaling,

05:09 a product can preserve its ideals.

05:12 If an investor or investment firm

05:17 needs a rapid return

05:20 or wants to make money quickly,

05:22 they may not allow the product

05:25 to move at its original pace

05:26 or follow its original direction.

05:28 That could cause

05:30 the product's values to drift,

05:34 and it could also change

05:38 what the founder is willing to stand by.

05:44 I imagine history should have

05:48 many cases where investment—

05:52 where investors or investment firms

05:58 were too eager to recoup their money.

06:05 Under that pressure,

06:07 the product

06:10 was forced down a different path,

06:13 ultimately leading to its demise.

06:19 There must be many examples like this.

06:23 But it seems these examples usually

06:26 remain largely outside our field of view.

06:32 If something succeeds,

06:35 everyone hears about it.

06:36 But when it fails,

06:39 most cases remain invisible,

06:42 without being seen.

06:49 Returning to "Tech for Good,"

06:52 another direction occurred to me:

06:55 if an ordinary person

06:58 wants to contribute

07:00 to Tech for Good,

07:03 what can they do?

07:05 I immediately thought of open source.

07:08 When I first learned about it, wow—

07:10 I thought the idea was wonderful.

07:14 I was producing a podcast for a VC firm,

07:20 and the episode was about

07:21 open-source communities.

07:23 Everything the guest described

07:26 about the open-source world

07:28 made me think, wow,

07:29 this is fascinating.

07:31 So interesting.

07:32 Later,

07:33 after learning about AI coding,

07:36 I officially became

07:38 someone on the fringes

07:44 of the open-source world—

07:47 so marginal that no one could see me.

07:51 I developed a small tool myself

07:56 and open-sourced it on GitHub.

07:59 Probably not many people use it,

08:02 but at least I've made contact with this world.

08:05 I still find that quite amazing.

08:08 So I asked AI,

08:10 "Is open source a form of Tech for Good?"

08:14 AI told me

08:16 that for the open-source contributor,

08:20 choosing to share technical understanding

08:22 and implementation

08:23 can be a way of doing good.

08:27 But it may not ultimately lead to good.

08:32 Another company or organization

08:36 might use your open-source work

08:38 to create a future where tech does not serve the good.

08:44 It's a little like

08:48 someone using what you open-sourced

08:50 and then

08:51 using it to do

08:52 the opposite of open source.

08:55 That's one possible interpretation.

08:58 It also said

09:00 Tech for Good goes beyond my examples,

09:04 such as the CEO's values,

09:06 the investors' values,

09:08 and also...

09:10 um...

09:12 the product itself.

09:13 That's what I mentioned earlier:

09:15 the product's use of data

09:17 and how it guides its users.

09:21 So there are roughly three dimensions.

09:25 The first is values,

09:27 whether founder, team, or investors.

09:31 The second is capital.

09:33 This includes firms and investors:

09:37 their expectations for returns

09:41 and their...

09:46 financial patience, perhaps?

09:48 Some need

09:51 to recoup their money quickly.

09:53 Others believe in long-term value,

09:56 so they give the team more time.

10:00 Then there's the product itself:

10:02 the product's data practices,

10:06 how it educates users,

10:08 and a whole lot of other things.

10:10 Those are roughly the three dimensions.

10:12 These three dimensions...

10:16 How should I put it?

10:20 At a surface level,

10:22 they do affect the goal of Tech for Good.

10:27 But I don't really know.

10:28 I'm not actually

10:30 a member of this field myself.

10:34 As an ordinary person,

10:37 all I can do is share technical understanding

10:40 and implementations.

10:44 That seems to be all I can do.

10:48 But returning to

10:49 the idea of tech-enabled poverty relief,

10:51 I truly want to do this kind of work.

10:54 But

10:56 I can't place myself

11:00 in an environment starved of resources.

11:06 So when I see posts on Xiaohongshu

11:12 about AI agents for education,

11:19 specifically to help

11:21 teachers in rural areas prepare lessons,

11:25 or about the work of edtech companies,

11:33 I genuinely think

11:36 they're doing something truly great.

11:39 So I'm also trying

11:41 to help them however I can.

11:43 I connect them with former classmates

11:50 who teach in small towns,

11:54 so they can test the products

11:57 and provide feedback.

12:01 But honestly,

12:03 although I've left frontline education,

12:05 I've realized that my extended family

12:08 still includes quite a few teachers.

12:11 Some teach in small towns,

12:14 and others in city centers.

12:20 That's roughly what I know.

12:25 So, returning to Tech for Good,

12:30 someone does need to work on its values.

12:36 And of course,

12:38 someone also needs to work

12:40 on tech-enabled poverty relief.

12:43 In places we're unaware of,

12:45 many people

12:46 still can't

12:48 make good use of a smartphone

12:52 to create opportunities for growth.

12:57 For example, when I visit my maternal grandmother,

13:02 her home is...

13:04 She lives in Hechi,

13:08 a remote mountainous place.

13:12 Honestly, wow.

13:13 I used to think—

13:17 before I visited her,

13:19 I thought my mom suffered with my dad.

13:24 But after visiting my grandma,

13:26 I realized my mom had a good life with my dad.

13:31 You know that kind of

13:32 complete reversal in perspective?

13:36 Oh...

13:37 It's so difficult.

13:38 I think people in Hechi,

13:42 especially those living in the mountains,

13:44 really live...

13:46 Perhaps they feel happy and content.

13:49 But maybe because

13:51 I've gone out and seen the world,

13:53 I feel

13:55 that living there would be so hard for me.

14:00 I mean...

14:03 I can't...

14:06 How should I put it?

14:07 Sometimes I want to help others,

14:12 but that savior complex isn't necessary.

14:16 Some people don't need to change,

14:18 and don't need you to bring anything new.

14:21 It reminds me—

14:25 I once heard a podcast

14:27 about volunteer teachers.

14:30 I forget whether the guest said this,

14:35 but some mountain communities

14:39 don't actually want volunteer teachers.

14:42 Especially now,

14:43 many universities run such programs.

14:47 Some remote communities

14:49 don't want volunteer teachers to come.

14:52 Why?

14:53 Because

14:54 these teachers bring us

14:59 so many advanced ideas.

15:01 It's as though they

15:03 open a window for us onto the outside world.

15:05 But many people lack the means

15:08 and opportunity to leave that place.

15:11 So what you bring them

15:12 becomes a source of pain.

15:15 Once someone's mind has been awakened,

15:18 if they lack the means

15:19 and opportunity to leave,

15:24 it can be

15:26 extremely, deeply painful for them.

15:30 When I heard this perspective

15:32 on that podcast,

15:34 it felt like a jolt to my brain.

15:40 I was genuinely shocked.

15:42 I'd never considered these questions.

15:44 I had assumed that bringing

15:46 new things to them was good,

15:49 and showing them a larger world was good.

15:51 But I had forgotten about

15:55 their first-person perspective—

15:57 the people who actually live there—

16:03 and what they feel.

16:06 Many of us think we're doing good,

16:09 that we're good people.

16:11 But have we truly considered those people?

16:15 Have we considered

16:16 whether they want to accept it?

16:19 Even if they do,

16:21 if they ultimately

16:22 can't leave that environment,

16:29 will it make them happier

16:31 or less happy?

16:33 I think that's also an important question.

16:38 So, returning to Tech for Good,

16:41 defining values for AI agents

16:45 is indeed an urgent task.

16:47 What about technology-enabled poverty relief?

16:54 I think it should be

16:59 something deeply meaningful.

17:02 When people learn how to make better use

17:05 of the technology they already have,

17:08 I think that should make them happy.

17:13 But then another question arises:

17:17 do you assume everyone

17:18 needs to grow in the same way you do?