
Co-Founder & Designer @KuberaDotcom | Ex Product & Design @newtonmailapp
Sydney via MEL, BLR, MAA, COK
Joined December 2006
- Tweets15.8K
- Following874
- Followers2K
- Likes11.5K
Umesh Theratt retweeted
Hi, it's Darshita. I am a designer.
And I'm pitching myself in public 'cause I'm done with a year full of anxiety.
Everyone says design is dead and AI killed it. But the real parasite was hiding in plain sight.
So here's my film, DESIGN IS DEAD. She's bold, fearless and very PINK 💖
Hiring? My DMs are open.
Umesh Theratt retweeted
The star of the day for me was this woman, who stood with the national flag held high for over 5 hours in CP. Listen to her articulate her reasons for doing so. Just amazing.
For decades, software followed a familiar model.
There was an operating system. There were apps. They barely knew each other existed.
AI could finally conduct the whole thing.
medium.com/@ooomz/reimaginin…
Umesh Theratt retweeted
Software is increasingly treated as disposable in the AI era. Yet we spend half our waking lives looking at screens! We don’t want to spend that time with zombie products—monotonous, vacant, and uncared for. We want products with soul. 💜
I go deep on this in my talk from @lennysan's Summit: youtube.com/watch?v=GLvFTMtw…
十年前,一个MIT大一新生,在冰箱里装了一个摄像头。
他怀疑室友在偷吃他的东西。
计划很工程师:拍下来,写个AI,当场抓获。
小偷没抓到。
他先被自己拍下的海量视频淹没了。
那一刻他想明白了一件事:
AI 缺的不是算法,是数据。
要做成它,得有现在一百万倍的数据。而这堵墙,每个做AI的人都会撞上。
这个新生叫 Alexandr Wang。
他出生在新墨西哥州的洛斯阿拉莫斯,美国第一颗原子弹诞生的小镇。
父母是中国移民,都是镇上国家实验室的物理学家。
六年级,他第一次参加数学竞赛,目标是赢一趟迪士尼之旅。
他赢了。
后来一路打进美国数学奥赛集训营、美国物理国家队。
17岁,他一个人去了硅谷,在旧金山四处借住沙发,挨个找CEO自荐。
简历几乎是空的,他只说了一句:我热爱写代码,我会跑得飞快。
Quora 创始人 Adam D'Angelo 收下了他。
在 Quora 干了两年,他才去读 MIT——
然后,就有了冰箱里那个摄像头。
2016年夏天,19岁的他和搭档 Lucy Guo 进了 Y Combinator。
两人在一间泳池小屋里办公,晚上睡气垫床。
他跟父母说,这只是个暑期项目。
然后,再也没回学校。
公司叫 Scale AI,干的是当时AI圈最不性感的活:数据标注。
别人在研究改变世界的模型。
他在雇人给图片画框,告诉自动驾驶汽车:这是个行人,别撞。
然后,风口一个接一个砸了下来。
先是自动驾驶,Waymo、Cruise 都成了客户。
再是美国空军和陆军。
2021年,Scale 估值 73亿美元。
24岁的他,成了全球最年轻的白手起家亿万富翁。
一年多后,ChatGPT 横空出世,大模型训练离不开人类反馈数据。
OpenAI、微软,都在用 Scale 的数据。
2025年4月,Meta 发布 Llama 4,口碑崩了,还陷入刷榜争议。
据杨立昆后来透露,扎克伯格非常恼火,对参与项目的所有人都失去了信心。
两个月后,扎克伯格出手:
143亿美元,买下 Scale 49% 的股份。
附加条件:Wang 本人必须来 Meta。
28岁的他,成了 Meta 史上第一位首席AI官,执掌新成立的超级智能实验室。
扎克伯格在内部信里,称他是"这一代最令人印象深刻的创始人"。
Wang 在给 Scale 员工的告别信里写:这一次的代价,是我的离开。
这个决定,难在哪里?
Meta 的AI阵营里,本来坐着一位图灵奖得主,杨立昆,深度学习三巨头之一,在 Meta 待了整整12年。
重组之后,他要向这个大学没读完的28岁年轻人汇报。
2025年11月,杨立昆宣布离开,出去创业。
后来接受《金融时报》采访,他这样评价这位年轻的上司:
"没经验。"
他承认对方学得很快,但说他根本不懂研究是怎么做出来的。
质疑声从来没停过。
媒体爆出,他在会上和 Meta 的老将们意见不合。
2026年3月,Meta 另起了一个AI工程团队,不归他管。
有媒体直接起标题:扎克伯格可能已经"受够了"他。
Meta 出面辟谣,扎克伯格在 Threads 上晒出一张两人挽着胳膊的合影。
一个月后,超级智能实验室交出第一份答卷:Muse Spark。九个月,把 Meta 的整套AI技术栈推倒重来。
有评论说,Meta 总算重新回到了牌桌上。
2026年9月8日,Meta 发布个人AI智能体 Muse。
订行程、回邮件、填表格,甚至替你讨价还价。
它冲上了苹果 App Store 免费榜第一。
第三方数据显示,它上线头12天的日活,是 ChatGPT 当年同期的将近3倍。
9月23日的 Meta Connect 大会上,Wang 宣布:Walmart、Best Buy、Gap、Sephora、Wayfair,都在接入 Muse。
十年前,他雇人给图片画框。
现在,他的AI开始替人下单。
杨立昆说他没经验。
这话没说错。
但 Wang 对此,有另一套说法:
"天真,是有巨大溢价的。"
不知道一件事有多难,才敢一头扎进去。
专家脑子里装着一张"此路不通"的清单。
创始人手里没有这张清单,所以会先去把路走一遍。
扎克伯格押的,就是这份溢价。
杨立昆也没有输。
他带着"大模型走不到AGI"的判断,拿到了超过 10亿美元融资,去做自己坚信的世界模型。
这从来不是谁对谁错的故事,是两条路线,各自押上了全部筹码。
赌局还没结束。
但扎克伯格这一注,第一张牌已经翻开了。
是张好牌。
Umesh Theratt retweeted
35 years ago I did it all myself. Wrote, built, sold, and released software solo.
25 years ago I did all the design myself. Sketches, then HTML, then CSS.
15 years ago I still did some hands on design, but mostly directed others to implement my vision.
5 years ago I primarily directed others. My "I can do it myself" muscles had atrophied.
Today, I feel like I'm back to the beginning. Able to write and design software entirely by myself. And having an elite team of others who can do the same thing is an enormous multipler.
Wild how massive advancements can propel you FORWARD by bringing BACK capabilities you lost along the way.
Full circle feels more like an arrow.
Umesh Theratt retweeted
Muse on Mac can now take over your computer and do things for you! Local computer use has shipped at Connect! 🥳
This is what I worked on at my startup 2 years ago with many great folks before joining Meta. Very excited to see it now shipped for the masses!
Umesh Theratt retweeted
$META unveiled Muse Charm, a palm-sized 5G device built specifically for talking to its Muse AI agent.
They plan to ship the device in December, in time for the holidays.
Sometimes I want to explore places, hotels on a map, look at photos, compare rooms and choose.
Other times I just want to say, “Find me the best hotel near Shinjuku under $500.”
The first needs a great UI. The second needs an agent.
The best software will know what needs a UI and what’s better left to AI.
If it’s frequent, visual, or worth exploring, give it a UI.
If it takes 12 clicks and a help article, let AI do it.
The interface stays simple. The product gets far more capable.
Umesh Theratt retweeted
Jev, a new kind of AI model, is showing developers a cheaper and faster path to software intelligence.
spr.ly/6018BGXyam
Umesh Theratt retweeted
I just read what Zuck wrote about Muse connectors.
A few things I think this means:
1. We’re witnessing the agentification of consumer apps.
2. Zuck believes Muse reaching 100M+ users feels entirely possible.
3. Whoever you connect to becomes your new landlord, so pick carefully.
4. Connectors become the new app listings (valuable real estate).
5. Being early could be as valuable as being early to the App Store in 2009.
6. Meta sees what people want, which providers convert, and what users will pay. That gives it the power to rank connectors, charge for distribution, and launch competing services.
7. The opportunity map includes agent native APIs, connector agencies, SEO for agents, identity and reliability infrastructure, and vertical connector marketplaces.
8. We’re moving from humans choosing apps to agents choosing businesses.
9. Every business will need an agent strategy, just like every business needed a mobile strategy.
10. Kinda crazy to say but a 3 person API company could reach 10M+ users through one great connector.
11. Connectors could create an entirely new second-order economy. The biggest opportunity may be helping businesses distribute, price, and optimize their services for agents.
12. Your API becomes more important than your app. Weird to think about but true.
13. Zuck wants Muse to own the moment between wanting something and getting it. He is coming for the app economy. Will he win? Right now it's Instinct vs Muse vs Grokbot. TBD.
Opening access for developers to build Muse connectors. You bring the API -- Muse brings the agent, the browser, and the context of what the person actually wants. People reach your service just by asking for it, and their agent takes it from there.
New connectors are live today. Come build with us. muse.ai/platform
Umesh Theratt retweeted
"I'll never look back. Love it!!" reviews.io/company-reviews/s…
this is the moment. 9 years in the making. what we've been building toward since 2017.
today we rolled out Matic Cues.
Matic is now the first home robot to understand 70+ spoken languages and human body language, including pointing at a mess on the floor.
humans have understood each other through cues for 200,000 years.
academia spent decades on human-to-robot interaction. we think the answer was always simpler: the best way to talk to a robot is the same way we talk to each other.
this matters most for families. a family isn't just app-savvy adults. it's kids. grandparents. people who don't speak English. people who just don't want another app.
Cues means anyone can use Matic.
today is the beginning of the end of the app.
for everyday cleaning, you won't need it at all. you'll just say:
"hey Matic, clean this."
"hey Matic, clean the kitchen."
"come here."
"follow me."
"go to the living room."
"come to the sink."
in any of 70+ languages.
the way you'd talk to anyone.
Introducing Cues: Voice & Gesture control for Matic
We raised $115M and spent 9 years to make the world’s first intuitive home robot
Say you spilled coffee: Point to the spill and say, "Hey Matic, clean this" and it will hear, see, locate in 3D, and go clean on its own.
Matic comes with a lot of features:
1. Say "Hey Matic" - it locates your voice, turns, and looks at you
2. Say "Hey Matic, follow me" and start walking. Matic will follow behind
3. Say "Hey Matic, go clean the living room". Since it knows your house map, it navigates and just does it
It's so easy, a 5 year old and an 80 year old can use it and it understands 75 different languages.
Matic has 8x the airflow, specialised cleaning algorithms for rugs, corners, toekicks, mopping, etc and cleans better than any other robot vacuum.
Also keeps improving with software updates.
13,000 families use and love Matic.
WIRED magazine gave it a 10/10 (the only hardware to receive this rating in a decade)
Buy yours at maticrobots.com and if you don't love it after 6 months, we'll give you a full refund.
To celebrate our launch, we're cleaning 300 homes with Matic in San Francisco and New York City.
Comment "Matic" below, we'll send you the link to sign up and come to your doorstep to clean your home with Matic.
Umesh Theratt retweeted
Got to spend some time with the Audi Novulari yesterday and I absolutely loved it. The proportions and details, especially in person, were just wonderful. Bravo @Audi.
Umesh Theratt retweeted
Dan emailed...
---
My name is Dan and I am an avid listener to Rework as well as a big fan of Basecamp, HEY and Omarchy.
I am at the early stages of building my own business and would love your thoughts on building a product in a competitive market? In my space, there are competitors but I am confident I can build something better. I wonder how much better does it need to be.
As 37signals have built products competing with giants like Jira, Gmail and Apple, I'm curious if that scares you? Do you worry about the big players?
---
I replied...
---
Hey Dan-
Every market is competitive. Either right now or later if you're doing well.
Focus on your own thing. Do your best work. You can't worry, think, or concern yourself with what other people are/can/might do. Do you.
The absolute worst thing is to pay such close attention to what everyone else is doing that you trick yourself into just building what they're built. Don't do that, or you don't need to exist.
---




















