ฝึกพูดภาษาอังกฤษด้วยเทคนิค Shadowing จากวิดีโอ: How to Build an AI-Native Services Company
กำลังสร้างบทเรียน...
1
Some of the biggest companies of the next decade won't be software businesses at all.
2
They'll be services companies like insurance carriers and law firms.
3
rebuilt from scratch with AI doing most of the work.
4
These are what we call AI native service companies.
5
And the markets are trillions of dollars in size: tax, audit, insurance, law, parts of healthcare, and so forth.
6
This opportunity didn't exist even a couple years ago.
7
But advances in the models have unlocked this new type of business, where companies provide the outcome to the customer, versus build a co -pilot that the customer uses internally.
8
These companies also look and feel different than most startups today.
9
In this video, I'll walk through a playbook for founders starting AI services businesses from scratch.
10
It's aimed at people thinking about starting a company, not if you're already running one.
11
I'll share some obvious and non -obvious elements of building these businesses that we've observed here at YC.
12
Topics include: picking a market, forming a team, building the actual product, serving the customers, the P &L,
13
and whether or not you should even buy a business.
14
One general comment before we get started.
15
We're still early here.
16
Like most things in AI, the market is moving fast.
17
We're learning as we go.
18
But the early successes here should get you really excited.
19
First, Picking the right market.
20
The same general advice for all startups applies here, with some important caveats.
21
You should pick a market you're excited to work in for a long time.
22
These companies still take a decade or more, If you don't love some combination of the customers or the market or the technical problem, you're not going to make it.
23
And that part isn't really new, but the best markets for AI services have four new, pretty unique traits.
24
The first is low trust.
25
Meaning the work is already outsourced and the customer cares about the final product, not how they got there.
26
you're displacing a vendor not asking the customer to do something fundamentally different That's a huge deal because you're not changing behavior.
27
You're showing up where the budget already lives and doing the work.
28
Second, low judgment at the task level.
29
If you can break the work into pieces and every piece needs a human exercising actual judgment, you can't really scale.
30
You need most of the steps to be automatable with judgment focused in a few places where humans stay in the loop.
31
The third is a high intelligence threshold.
32
This sounds contradictory, but actually it isn't.
33
The overall work has to be hard.
34
Hard enough that models plus humans are needed to actually deliver an outcome the customer accepts.
35
The fourth is: Regulation could actually be good.
36
Regulated industries have higher expectations and legal accountability.
37
That raises the bar and the moat for founders.
38
For instance, Panacea is a current YC company that provides FDA regulatory services for biotechs and medtechs.
39
They actually hire experienced FDA consultants, pair them with an AI platform to deliver faster, higher quality FDA approvals.
40
So what are some of the specific markets we have in mind here at YC?
41
The known good fit markets include tax, audit, insurance, mortgages, parts of healthcare, and parts of logistics.
42
But there are plenty more markets nobody has touched yet. about on X.
43
And here are a few more things to keep an eye on when it comes to the markets.
44
The first is on the models.
45
Will the models disrupt these businesses?
46
It depends on what I call the Sam Altman test.
47
You should ask yourself, as the models get better, does your service get stronger, Or does the model itself commoditize you?
48
You want to be in the first camp.
49
Where to be careful?
50
Anything involving equipment and on -site labor.
51
The software margin math doesn't apply when you own and operate physical things.
52
It's very hard to create real leverage, though these can be really good businesses.
53
Let's leave this area to the robotics founders.
54
One more honesty check.
55
Ask yourself sincerely, are you using humans because the work genuinely needs judgment, or are you compensating for product gaps?
56
Be honest here so you're not papering over product shortcomings with actual humans.
57
There are still great, massive technology businesses to be built with humans in the loop.
58
Second, and maybe most importantly, the right founding team.
59
The same advice applies for all startups here, but again with some important caveats.
60
You should build companies with people you already know and you've worked with.
61
If you're solo, think about the best people you've ever worked with and ask them to join you.
62
You'd be surprised who says yes.
63
For AI services specifically, there's three attributes that all the best founders share.
64
The first is domain fluency.
65
Direct experience is best, but learned is actually okay.
66
You're selling to skeptical buyers in often regulated spaces.
67
You have to bleed credibility.
68
How you acquire it matters
69
somewhat less the second is model fluency you need to know what frontier models can do today
70
and design the product to ride the curve as they get better there is no substitute for great tech here
71
People underestimate this.
72
Next is operational rigor.
73
Topics like variance, throughput, cycle times, SOPs, this is not an exciting set of words for most founders.
74
but you are fundamentally running an operation.
75
You have to learn that skill set and you have to enjoy it, or at least you have to respect it.
76
The product is in operation.
77
A great example here is the General Legal Team, which is an AI -native law firm that YC recently backed.
78
The founders have a unique mix of actual law firm experience at Cooley and Fenwick, as well as years of technical leadership at Case Text.
79
But most importantly, they think deeply about throughput and how they staff their firm.
80
They've integrated shift work into how they serve clients to reduce cycle times and attract the best lawyers on the team.
81
This is a win -win for scale.
82
Now let's talk about building the actual product.
83
With AI services, the setup is the opposite of most software.
84
The human is the interface of the customer, not the product.
85
The product helps the humans scale their work non -linearly.
86
That changes pretty much everything around building an actual product.
87
First, you need to apply an operations mindset.
88
find the bottlenecks and build for the bottlenecks.
89
Throughput and cycle time are now product metrics, track them like you would daily active users.
90
Variance is the existential problem here.
91
By variance, I mean non -uniform outputs from your actual service.
92
Customers will fire you for variance faster than they will fire you for being a bit slower
93
or a bit more expensive than the incumbents.
94
They need to trust the output.
95
Inconsistency destroys trust, which causes churn.
96
Thirdly, humans in the loop should scale non -linearly.
97
If revenue scale is just in line with the number of humans you add, you'll have major problems.
98
The humans in the loop also need to enjoy the software.
99
They are your users.
100
A general point: it's okay to do things that don't scale at the very beginning, but eventually you really do need to scale.
101
Automating the process is the product.
102
Okay, sales and customer success.
103
The biggest challenge facing founders here is what I'll call the early demand trap.
104
It's easy to sign up a lot of pilot customers when you're just starting out and have nothing, but it can quickly overwhelm your ability to serve them, and you won't be able to build the product to scale.
105
You'll be stuck using humans.
106
It is a literal trap.
107
Our advice here is to cap your first pilot customers to a small handful.
108
Resist the temptation to sign too many too quickly.
109
Assuming you avoid this early demand trap, pre and post sales looks pretty different too.
110
You have to sell outcomes, not seats or tokens.
111
The pilot is the product.
112
For the first handful of customers, don't try and standardize too early.
113
Use those pilots to learn.
114
Find the spots where AI gives you unique leverage versus spots where you're just automating something obvious.
115
Build the product accordingly and do it fast.
116
Pricing is harder than traditional software because you're not competing with other software providers.
117
You're competing directly with the cost of labor, internal or outsourced.
118
A few options here on pricing.
119
There's per unit pricing, so per return, per claim, per loan.
120
This is the cleanest, it's the easiest to explain.
121
There's also outcome based pricing.
122
This aligns incentives beautifully, but it can be harder for you to forecast in your business.
123
Panacea prices on the completed consultant study versus hourly, which is the norm in the industry.
124
There's definitely two pricing strategies to avoid.
125
Cost plus pricing caps your upside permanently don't do it.
126
Straight line undercutting makes your work seem cheap and potentially low quality.
127
"Price on value." Next, the P &L, or the profit and loss statement.
128
This is where these companies live or die, so let's do a quick walkthrough.
129
If you haven't stared at one of these financial statements before, you're not alone among founders, and that's okay.
130
The general structure is, Revenue minus cost of goods sold gets you to gross profit.
131
Gross profit minus operating expenses gets you to operating income.
132
Let's dive into each of these in the context of AI services.
133
Revenue.
134
Ah, the easy part, relatively speaking.
135
You will be able to sign contracts.
136
Can you deliver on them repeatedly?
137
I don't know.
138
That depends on your product and your process.
139
Eventually, you want smooth and predictable growth.
140
A great product process is going to smooth out that lumpiness, but in the early days, it'll be spiky on a monthly basis.
141
That's okay.
142
Cost of goods sold, or COGS, obsess over this from day one.
143
There's three main components here: there's model costs, hosting costs, and those humans in the loop.
144
All three of them need a number, a trendline, and someone who owns them.
145
Be deeply suspicious of zero margin or negative margin pilots.
146
They're fine to learn from, but it's really dangerous to get hooked on those.
147
The core bet here is, The more the product is built, the lower the cogs, the better the gross margin.
148
i call this ai operating leverage okay opex this includes research
149
and development costs so building the product sales and general administrative
150
which is like finance legal exec salaries pretty standard stuff here
151
operating income this is the revenue minus the cogs minus the
152
opex you will be judged on your operating income in these businesses faster than you might expect.
153
Net income.
154
This is the operating income less taxes and interest.
155
That's less important in the medium term.
156
So here's the actual P &L opportunity for these businesses.
157
Traditional services firms top out around 30 % margins.
158
Pure software and agent companies have more margin, but often smaller TAMs.
159
The bet on these services companies is that that AI operating leverage gets you closer to software margins, say 50 % plus, on a market that's two to three times bigger than software.
160
You don't need to be there right away, that's okay, but the trajectory has to be believable.
161
Okay, last but not least.
162
Don't try to buy your way in.
163
There's a temptation we've seen, especially among founders with some operating background, to try and buy an existing services business.
164
Add some AI on top, short circuit the revenue.
165
This is generally a trap.
166
There's one decent reason to do it you need a regulatory mode fast.
167
Insurance licensing, for example.
168
But otherwise, this almost never works.
169
So why is that?
170
You just can't acquire a product market fit.
171
Legacy service businesses are, you know, legacy.
172
They have different expectations on metrics, hiring, and performance.
173
Adding AI on top of that doesn't immediately change any of those realities.
174
Building is almost always better than buying.
175
Okay, to recap, AI services companies present an incredible opportunity for today's founders, but these are fundamentally different startups to build.
176
If you avoid the traps and focus on the process as the product and the product as the process, You have a chance to create a generational company.
177
We're excited to see what you build, and hope you apply to YC.
178
you
📺 ช่องเดียวกัน
C11
Garry Tan: Own Your Intelligence
Business EnglishUniversity StudentsWorking Adults
B22
Dot Plots: How to Actually See What Your Users Are Doing
Business EnglishUniversity StudentsWorking Adults
B25
How Pitching Investors is Different Than Pitching Customers - Michael Seibel
Business EnglishInterview EnglishUniversity Students
C14
Jensen Huang: The Mindset That Built NVIDIA
Business EnglishUniversity StudentsWorking Adults
B2How to Get Your First 10 Customers
Business EnglishWorking Adults
✨ วิดีโอแนะนำ
C11
126 ประโยค
How I use AI as a senior software engineer and 5 principles of AI explained in 10 minutes
Business EnglishUniversity StudentsWorking Adults
B21
264 ประโยค
Inside a YC Startup Disrupting a $900B Industry That Runs the US
Business EnglishUniversity StudentsWorking Adults
B21
236 ประโยค
how I built a $2.7M brand using a.i (my actual product, website, ads, viral videos)
Business EnglishUniversity StudentsWorking Adults
C12
39 ประโยค
The Global Power Play
Business EnglishUniversity StudentsWorking Adults
C13
460 ประโยค
Welcome to Agentic GTM: The AI Revolution in Sales
Business EnglishUniversity StudentsWorking Adults
C11
159 ประโยค
Top 5 AI Gateway Use Cases | Solo.io
Business EnglishUniversity StudentsWorking Adults
C11
409 ประโยค
I think AI slop is about to trigger a renaissance, actually
Business EnglishUniversity StudentsWorking Adults
C15
182 ประโยค
How to Become an AI Product Manager in 2026 | Ex-Google, Microsoft
Business EnglishUniversity StudentsWorking Adults
C11
103 ประโยค
Amazon Bedrock Tutorial | Getting Started with Amazon Bedrock | AI Model Access & Testing
Business EnglishUniversity StudentsWorking Adults
B21
201 ประโยค
Full Video: Anna Chapman on her startup Property Finder
Business EnglishInterview EnglishUniversity Students
เกี่ยวกับบทเรียนนี้
ในบทเรียนนี้ ผู้เรียนจะได้ฝึกพูดภาษาอังกฤษโดยการฟังและเลียนแบบเนื้อหาจากวิดีโอที่พูดถึงการสร้างบริษัทบริการที่อิง AI ผู้เรียนจะได้เรียนรู้เกี่ยวกับยุทธศาสตร์ต่าง ๆ ในการสร้างธุรกิจ รวมถึงการเลือกตลาด การสร้างทีม และการบรรลุเป้าหมายของผู้ใช้ ซึ่งเป็นเนื้อหาที่มีความสำคัญสำหรับผู้ที่สนใจในโลกของเทคโนโลยี AI
คำศัพท์และวลีที่สำคัญ
- AI native service companies - บริษัทบริการที่อิง AI
- low trust - ความไว้วางใจต่ำ
- automation - การทำงานอัตโนมัติ
- regulation - กฎระเบียบ
- high intelligence threshold - ระดับปัญญาสูง
- customer outcome - ผลลัพธ์สำหรับลูกค้า
- market fit - ความเหมาะสมของตลาด
- judgment - ความชอบธรรม
เคล็ดลับในการฝึก
เพื่อช่วยให้คุณปรับปรุงการออกเสียงภาษาอังกฤษได้ดียิ่งขึ้นในการฝึกฝนนี้ ขอแนะนำให้คุณใช้เทคนิค shadowspeak หรือ shadow speak โดยการฟังและเลียนแบบเสียงจากวิดีโออย่างใกล้ชิด ในขณะที่ฟังให้อยู่นิ่ง ๆ และพยายามให้เสียงของคุณตรงตามน้ำเสียงและจังหวะของผู้พูด
ระหว่างการฝึก ควรจับจังหวะของการพูดให้ได้อย่างถูกต้อง พยายามเน้นคำสำคัญให้ชัดเจน และปรับระดับเสียงให้สอดคล้องกับน้ำเสียงของผู้พูด เริ่มจากการฟังซ้ำ ๆ และพยายามพูดตามให้เร็วขึ้นเรื่อย ๆ โดยใช้บริบทจากเนื้อหาในวิดีโอ ฝึกพูดบ่อย ๆ ด้วย วิธีนี้จะช่วยให้การออกเสียงของคุณพัฒนาและช่วยให้คุณสามารถใช้ shadowing site เพื่อสะสมประสบการณ์ในการพูดอังกฤษของคุณได้
ไวยากรณ์ในวิดีโอนี้
โครงสร้างที่ผู้พูดใช้บ่อยที่สุด พร้อมคำพูดจริงจากวิดีโอ:
| โครงสร้าง | ในวิดีโอ |
|---|---|
| Present perfect have/has + กริยาช่อง 3 — เหตุการณ์ในอดีตที่ยังเกี่ยวข้องกับปัจจุบัน | have unlocked · we've observed · has touched |
| Passive voice be + กริยาช่อง 3 — เน้นสิ่งที่เกิดขึ้น ไม่ใช่ผู้กระทำ | is already outsourced · are needed · be built |
| Relative clauses who / which + อนุประโยค — ข้อมูลเพิ่มเติมเกี่ยวกับคนหรือสิ่งของ | Team, which is · hourly, which is · someone who owns |
เทคนิค Shadowing คืออะไร?
Shadowing เป็นเทคนิคการเรียนรู้ภาษาที่ได้รับการรับรองทางวิทยาศาสตร์ พัฒนาขึ้นสำหรับการฝึกนักแปลมืออาชีพ วิธีการนี้เรียบง่ายแต่ทรงพลัง: คุณฟังเสียงภาษาอังกฤษจากเจ้าของภาษาและพูดตามทันที — เหมือนเงาที่ตามผู้พูดด้วยช่วงเวลาห่าง 1-2 วินาที การวิจัยแสดงว่าเทคนิคนี้ปรับปรุงความแม่นยำในการออกเสียง ทำนองเสียง จังหวะ การเชื่อมเสียง การฟังเข้าใจ และความคล่องแคล่วในการพูดได้อย่างมีนัยสำคัญ