Mayank (@krmayank) header
Mayank (@krmayank) profile picture

Defining the AI-native infrastructure layer—how intelligent systems should be built, operated, and evolved safely at planetary scale. Ex Salesforce, Microsoft

Silicon Valley
Joined June 2009
Photo by Mayank (@krmayank)Photo by Mayank (@krmayank)Photo by Mayank (@krmayank)Photo by Mayank (@krmayank)Photo by Mayank (@krmayank)Photo by Mayank (@krmayank)Photo by Mayank (@krmayank)Photo by Mayank (@krmayank)Photo by Mayank (@krmayank)Photo by Mayank (@krmayank)
Replying to @GShaneMorris
It’s the exact same use case as launching the old roadster into solar orbit near Mars. It’s sheer audacity. It is an example of what we can do, and more than that, what we can _afford_ to do. It is youthful exuberance and power in full display. It is a message to the rest of the world: “You can’t catch us.” Cue: We are the champions.
24
24
13
976
20,012
Technical moats are mostly over at this point. If you’re building something that is meant to be durable, the only moats now are: • Proprietary data • Network effects • A physical component
1,145
2,080
907
25,717
3,342,900
We are making Markdown files work natively across Google Drive and Docs. You can now preview .md files in Drive, open them in Docs, edit them, comment on them, and collaborate without converting the file into a Google Doc!
937
1,342
952
21,016
1,447,267
Mayank retweeted
Replying to @PSkinnerTech
yeah i feel like there are a lot of parallels between systems that provide constraints (eg type systems) and using agents effectively and at scale. when you are dealing with extreme volume that you cannot control directly you have no choice but to shape the environment. i was inspired by my own journey trying to use typescript effectively no.lol/2019-12-27-type-narro…
8
7
2
229
16,699
Cool eval. Simply ask an LLM “Land or Water?” and give it a latitude and longitude coordinate as text. Ask 16,200 times, plot as image. The models know. From compressing the internet.
Replying to @celestepoasts
results for all claudes
727
1,099
184
19,590
1,635,206
Mayank retweeted
here's how i shipped 2,500 PRs last month to production this was originally supposed to be for Cursor Compile in London. i couldn't make it since i was livestreaming for Grok @Bot Galaxy so i'm making it available for free here on X! watch it on 2x speed, i talk slowly
852
1,963
659
22,549
4,738,146
Here’s to the wanting. Here’s to the dreaming. Here’s to everyone.
808
1,207
544
14,089
5,985,076
The claude folks should just own this 🤣
221
600
263
4,017
704,374
There are two approaches I'm using with Jev, LLMs, and MCP: - Jev as a LLM Tool orchestrating MCP tool calls - Jev as the outer loop orchestrator w/ an LLM as an tool MCP output schemas are essential for giving Jev everything it needs to efficiently plan. Then we tell Jev about the tools and have it build a workflow AST.
15
36
3
296
14,469
New in Claude Code: your sessions can now message each other. Instead of having to re-explain yourself in another session, you can now tell Claude to do it. It sends a summary (not your history or files), and the other session picks it up mid-task.
1,898
4,573
2,474
59,745
7,991,296
Mayank retweeted
BLACKROCK CEO LARRY FINK SAYS A NEW ASSET CLASS IS COMING: FUTURES ON COMPUTE "The United States is short power. We're short compute, we're short chips." Fink sees shortages ahead across all three, plus memory. And he thinks the scarcity gets financialized: "I actually believe a new asset class will be buying futures of compute." That is computing power described the way traders treat oil: a commodity scarce enough to hedge, price and trade. "We just don't have enough compute power right now."
89
166
53
1,108
195,682
Mayank retweeted
Finally, sudo make me a sandwich is reality.
Today we're opening up the DoorDash CLI in limited beta. `dd-cli` lets you order DoorDash directly from your agent: search stores, find the best deals, check out, and more. Early access for US/Canadian macOS developers by waitlist. Excited to see what folks build!
124
353
22
5,966
627,904
My brother posts this every year in the family chat on the 4th. Still hilarious.
842
14,911
954
103,000
6,457,401
The first nail in the coffin of every sandbox-only startup:
61
90
51
1,326
429,385
40% of fatherhood is walking around the house, turning off lights.
431
1,597
316
20,096
3,382,446
Mayank retweeted
This is the most inspiring positive-sum vision for AI in the enterprise.
92
166
2
3,175
1,005,000
Mayank retweeted
🚨 Musk vs OpenAI's lawyer — the cross-examination exchanges William Savitt — Wachtell Lipton's lead defense lawyer, Supreme Court clerk, trained to break witnesses. Savitt opens with a misleading premise. Musk: "You're being misleading. What you're saying is false." Savitt tries again with a different loaded frame. Musk: "Your questions are not simple. They are designed to trick me." Savitt demands a yes or no answer to a complicated question. Musk: "If you ask a question where there is no possible simple answer, I must give a longer answer because any simple answer would be misleading the jury." Musk reaches for an analogy: "The classic answer to a yes or no question is not so simple. For example, if you ask the question 'will you stop beating your wife?'..." Judge Gonzalez Rogers cuts him off: "No, we're not gonna go there." The courtroom laughs. Savitt apologizes for the question. Musk: "I find it funny you saying it wasn't an unfair question since you're only asking unfair questions." Savitt: "I'm doing my best." Musk: "That is not true." OpenAI's lawyer came to break Musk. Musk wasn't having it.
909
3,367
400
32,115
5,125,317
My job used to be writing code. Now it's marriage counseling for Codex and Claude Code.
99
217
24
2,982
123,814