Once the FDE Fills the Gaps in Enterprise AI Adoption, the Role Dies
One-person companies don't hold up, small teams do; FDE is not a lasting business, just a phase that fills the gaps in enterprise AI adoption — once the gaps are filled, the function dies.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
Model vendors' token margins are squeezing third-party agents
Active users of third-party agent apps may be down a third from their peak. The reason isn't complicated: model vendors have very high token margins, so they can supply tokens at very low cost, while third parties buy tokens from them and can't fight a price war. More critically, the stronger the model, the smaller the difference a Harness makes — Manus used to pile on a lot of scaffolding to make it strong, and once the model upgraded, 80% of that scaffolding was dead. For ordinary tasks like writing slides, writing docs or doing research, a slightly better or worse Harness makes little difference.
OPC doesn't hold up — sell first, don't stockpile skills
A lot of unemployed people call themselves OPC and ask whether the most important thing is building a Harness or something else. The answer is none of it matters; what matters is finding who will pay you. Most OPCs are just pleasuring themselves, laughing at this and laughing at that, then conning a batch of people who also want to be OPC into giving them money — but nobody has money left, and a stream with no source can't last. Doing the work and turning the work into money are two different things; the hard part is the latter, and most of these people weren't in sales to begin with and have never sold anything directly in a market.
The information gap is bigger than you think: house calls to install Codex for bosses
A group of entrepreneurs think you're an AI expert, yet the questions they ask are how to install Codex and how to store a credit card. Someone specifically found a kid who came back from the UK, covering Shanghai for house calls and everywhere else by remote video, to swipe cards and download software for them. Many companies that look extremely high-tech have founders who normally use Doubao, and who ask Doubao about much of their work. Traditional enterprises run on thin margins — a VP might make only 500k, an R&D lead might make only 800k — and knowing the stuff isn't enough, you also need the ability to explain it, and the boss doesn't trust him anyway; the monk from outside is more likely to be believed.
The FDE's fate: once the gaps are filled, it dies
FDE may not be a lasting business; it's more like a phase that fills the gaps in enterprise AI adoption, and as those gaps slowly get filled, the FDE hand on the clock dies. Because the stronger the model and the more complete the product, the more it should be the company's own employees holding an agent and just spraying it themselves. What every company needs to solve isn't an AI problem, it's still a business problem — a restaurant needs customers to come and spend money eating its food. AI outsiders don't understand how a restaurant operates, restaurants don't understand AI, and if you want to make money you work desperately to understand the restaurant and slowly help it fill the gaps.
AI transformation is politics — the boss needs a knife from outside
Insiders know the company, so why bring in McKinsey to explain things? Much of the time it's politics. AI revolutionises yourself: for employees it's here to take their jobs, for management it's here to take their power, and operating on yourself is hard. The boss looks very authoritative, spraying subordinates like dogs, but he can no longer interfere in the business, and may not even dare fire that executive. The executive will tell the HR lead pushing AI transformation very directly what this does for him — the two know each other well, so he can't say there's no benefit. So you borrow a knife to kill: it's not that the position can't do it, it's that the position can't do it.
Distillation is an extremely high-ROI way to get data
Distillation is definitely useful, one of the highest-ROI things. The efficiency gap between having AI write code and having a person write code — that's the efficiency gap in preparing data. The way you get training samples is automatic, and as long as you have enough accounts and enough money it scales horizontally; whereas having people write data is manual and doesn't necessarily scale horizontally. From a post-training perspective, for getting bootstrap data, the efficiency difference between the two sides is enormous. As for why ByteDance hasn't produced SOTA, it's not necessarily related to distillation — Google hasn't either, and both ByteDance and Google do everything, so they may be aiming for a unified all-modality model, and early fusion makes the problem more complex.
DSH isn't a good Harness, but it's a good kernel
Good Harnesses or Agents all make a lot of trade-offs and have strong opinions: Codex has a plugin mechanism but you can't change its own Agent loop, Pi thinks you don't need many tools, OpenCloud values access through various IMs. DSH's approach is that everything can be changed, so it's definitely not a good Agent or a good traditional Harness, and people who spray it saying it's badly built are completely right. But its path has value: why is the whole coding harness a static thing, why can't this round of conversation be like this and that round like that, is the whole agent runtime also changing dynamically.
Wearable software: everyone needs their own kitchen
Overseas they call this Wearable Software. Two images make the point: one is a maker's studio, below it a kitchen — every chef's or electrician's workspace looks a little different, with different equipment, layouts, and the spatulas and seasonings they habitually use. That's how everyone uses an Agent too: some use Cmax, some use the Codex client, some use Herdr. This need used to be hard to satisfy, so everyone gritted their teeth and used a combination of different software, and customising in very small ways was hard; today writing code to customise to my needs has become less difficult, but you shouldn't build from scratch either — you should take some piece of software or kernel and change it into what you need.
AI-native org: let AI drive the work loop closed
What works better isn't changing the org so you use AI better, but using AI to perform the org's functions — AI is the org's competitor. Concretely: file a bug in Feishu, and the Agent automatically looks at online logs and the current codebase to answer, knowing who you are, your email, and the actual running state of your account; every day it scans online alerts and posts them to different groups; every week it automatically scans which modules are duplicated, which need refactoring, where docs haven't been updated and aligned, and AI sends you a PR with the changes. When a designer wants to change the UI, R&D hands them the codebase to change themselves, with a check gate added so they can only commit to a branch, and an engineer helps set up the environment so it can simulate the online server locally.
In their own words · checked verbatim
I think OPC itself doesn't hold up, but I think the team being small does hold up — I've always held this view. FDE may not be a lasting business; it's more like a phase that fills the gaps in enterprise AI adoption, and as those gaps slowly get filled, the FDE hand on the clock dies.
我觉得OPC本身不成立 但是我觉得团队归功小是成立的 我其实一直是这个观点 FDE可能它不是一个长久的业务 它更像在这个阶段去填补企业渗透AI缝隙的过程 随着慢慢填补了 FDE这个时针就消亡了
The most important thing is finding who will pay you. They'll say, for OPC the most important thing is building this Harness or whatever — I say it doesn't matter, what matters is the money.
最重要是找到谁给你付钱 他们就会说 OPC我们最重要是搭这个Harness 还是干嘛干嘛 我说不重要 不重要是给钱
Where are most people in the world stuck? Recently, by chance, I've been mixing with a group of entrepreneurs on some things. They think you're an AI expert, but the questions they ask me are all like how to install Codex and how to store a credit card.
世界上绝大部分人的卡点卡在哪里呢 我最近机缘巧合反正有一群企业家 一起混一些事情 他们觉得你是AI专家嘛 但是他们问我的问题 都是类似于Codex怎么安装 以及信用卡存储
What every company needs to solve isn't an AI problem, it's still a business problem. Whether I'm a retailer today, a restaurant today, or a manufacturing enterprise today, the problem I need to solve isn't how efficient my AI is, but that my restaurant needs customers to come and spend money eating my food.
所有的公司要解决的并不是AI问题 还是业务问题 我今天是个零售商也好 我今天是一个餐馆也好 我今天是一个制造业企业也好 我要解决的问题并不是我的AI多有效率 而是我这个餐馆得有客人来 花钱吃我这个麦
Because AI actually revolutionises yourself. For employees it's a little bit here to take their jobs, for management it's a little bit here to take their power — operating on yourself.
因为AI它其实革命的是自己 自己对于员工来讲 其实少少有一点事是来抢饭碗的 对于管理层多多少少 它有一点事来抢权的 让自己开刀自己这事
And the way you get training samples is automatic, and as long as you have enough accounts and enough money it scales horizontally. But having people write this data is manual, and doesn't necessarily scale horizontally.
而获取训练样本的方式是自动的 而且你只要有足够多的账号跟足够多的钱 它是横向扩展的 但是你要人去写这个数据 它是手动的 且不一定能横向扩展
But Deepseek's Agent feels like everything can be changed. I think an all-purpose thing is never equal — it definitely faces this problem: what is this thing actually for?
但是Deepseek的Agent的感觉就是所有东西都可以改 我觉得一个万能的东西万万不等的 就是他一定他面临这个问题 这东西到底该啥用呢
It's actually not changing the org so you use AI better, but using AI to perform the org's functions — AI is the org's competitor.
其实就不是改组织 让你更好的用AI 而是用AI来完成组织的职能 就是AI是组织的竞品
Figures
| Decline in active users of third-party agent apps | may be down a third from the peak | 3:55 |
| Manus cost per task | running one task burns $20 | 5:02 |
| Cloud Code subscription price | $200 | 5:02 |
| Traditional enterprise VP annual salary | maybe only 500k | 14:08 |
| Traditional enterprise R&D lead annual salary | at a not-small company, maybe only 800k | 14:08 |
| Kimi subscription price | 199 RMB | 23:11 |
| Grok subscription allowance | X Premium includes roughly $30 worth | 23:11 |
| Array microphone dev board price | about three or four hundred RMB | 1:02:28 |
| Camera price | maybe a hundred-odd RMB | 1:02:28 |
Glossary
- OPC / one-person company
- A company run by one person; the show argues it doesn't hold up, small teams do.
- FDE / forward-deployed engineer
- The role that goes on-site to help enterprises land AI; the show argues it dies once the gaps are filled.
- Harness
- The agent runtime wrapped around a model, deciding tools, loops and interception points.
- Wearable Software
- Software changed into what you need based on some kernel, rather than built from scratch.
- early fusion / late fusion
- Two decisions about when to add visual capability in multimodal training.
How to listen
Founders and engineers building AI applications, enterprise services and agent products, plus managers pushing AI transformation inside their companies.
The opening news roundup and the closing chit-chat about making videos and Jarvis can be fast-forwarded.