devangt (@devangt) header
devangt (@devangt) profile picture

🎨Art | 🧑‍💻Tech | 💸Economics | ⚽️Soccer (@arsenal & @soundersfc) | 🏏Cricket | Global Head of @christiesinc Ventures & Innovation Labs

Seattle, WA
Joined May 2008
Photo by devangt (@devangt)Photo by devangt (@devangt)Photo by devangt (@devangt)Photo by devangt (@devangt)Photo by devangt (@devangt)Photo by devangt (@devangt)Photo by devangt (@devangt)Photo by devangt (@devangt)Photo by devangt (@devangt)Photo by devangt (@devangt)
Replying to @theinformation
Inference at the edge of the telco stack. Clever Nvidia play.
97
Replit was building toward the AI coding era years before it had a name. In 2024, the company shipped an agent that could take a request in plain English, work through errors, provision its own infrastructure, and keep going until the software actually worked. The model companies even reached out to ask how Replit had pulled it off. Co-founder and CEO Amjad Masad had been thinking about making programming radically more accessible since 2008, and Replit’s original pitch deck predicted AI-generated code before the Transformer paper was published. Today, that same technology is moving into banks, manufacturers, and other large companies, where Amjad says teams are replacing software projects once scoped at $10 million with $250,000 of Replit. On How to Build the Future, @paulg sits down with @amasad to talk about why @Replit was so well positioned when coding agents arrived, what comes after vibe coding, and his bigger vision for the “self-driving company,” where AI agents increasingly build and operate the software a company needs. 00:00 - "You guys basically invented vibe coding" 02:43 - Watching users copy-paste from ChatGPT 05:27 - The origins of Replit 08:00 - People underestimate their ability to predict the future 10:27 - Grieving the craft 12:52 - A culture of obsessing over users 15:13 - Why enterprises pick Replit 17:28 - The agent was just another user 20:17 - It barely worked, and users loved it 23:13 - $10 million projects now for $250,000 25:45 - Sandboxing for humans 28:14 - "The self-driving company" 30:34 - The way AI flows into companies 33:28 - What Steve Jobs knew before the iPad 35:59 - Killing the sacred cow 38:25 - The rest of the world is two years behind 41:15 - The compute brought on AI 43:46 - The person who assumes dramatic change
90
56
18
388
58,333
Replying to @verge
AI research now has a feedback loop. The hard part is making safety compound as fast as capability.
18
Replying to @WIRED
Sometimes the best prompt is still a search query.
7
Replying to @Reuters
$11B at a $1.25T valuation. The bar is now revenue, not ambition.
20
Replying to @zerohedge
When the memory squeeze is bad enough to tempt TSMC, that tells you how tight it is.
4
556
Replying to @Reuters
When Microsoft starts trimming rented AI capacity, every neocloud should be paying attention.
265
Replying to @seattletimes
Prime keeps becoming the front door for every other app. Smart wedge.
317
Replying to @TechCrunch
Usage up, spend flat. The efficiency era just started.
1
2
144
Dev life before agents Dev life after agents
166
134
67
1,927
162,563
devangt retweeted
Replit is the only collaborative AI coding system out there.
I want Claude Code to: - Be a chat for my team - Every thread is a chat room we're both in - We can both prompt and steer a thread, or talk to each other and the agent can just keep cranking until it thinks we're saying something it needs to do differently Does this exist?
93
15
4
263
27,670
Replying to @Yuchenj_UW
948 kernel bugs from one model. Security budgets are about to get rewritten.
20
Replying to @seattletimes
Climate Pledge is ready. Seattle has waited long enough. Bring them home.
1
1
49
Replying to @dougboneparth
Basic intelligence
13
Replying to @business
Hardware out, platforms in. The AI trade is rotating, not ending.
71
I think MAAG7 should do capacity auctions for open contracts. I’d be happy to help Monet / Rothko = ( Blackwell + Vera Rubin ) x 10000
1
3
65
In $MSFT , $AAPL, $AMZN and $GOOG - 7 we trust.
5
165
Better margin than software
Dylan Patel says Elon is selling xAI's compute to Anthropic and Google because a GPU rents for $15 an hour and costs him $4 "He's raising another $40 billion right now. At the same time, he claims he wants to build 13 gigawatt. 13 gigawatts, if you build all of that out, is, you know, 400 plus billion dollars, right? You know, 500 plus billion dollars. So where does he get the cash, right?" "Being able to sell GPUs for, you know, $15 an hour when the TCO of them is only like $4 an hour, you have this massive delta. And so he's exploiting that." "Step two is, okay, how do I use that compute most effectively to dominate the universe, right? ... I'm sure it's just to benefit humanity." [ Jordi: A little bit. ] "If that means, hey, he ripped some compute out of the xAI's team to sell it to Anthropic and Google, he had to do that, right?"
3
258
Replying to @nikitabier
Yes did a big merge today and made hand edits — I kinda still love it — one of the PRs was Claude’s
1
221