Ankur Jain (@ankur10) header
Ankur Jain (@ankur10) profile picture

Stanford Alum, MIT Media Lab | Jivi AI BharatPe. @Instalocate. @WalmartLabs. Kosmix | Creating digital products since age 10

New Delhi, India
Joined March 2007
Photo by Ankur Jain (@ankur10)Photo by Ankur Jain (@ankur10)Photo by Ankur Jain (@ankur10)Photo by Ankur Jain (@ankur10)Photo by Ankur Jain (@ankur10)Photo by Ankur Jain (@ankur10)Photo by Ankur Jain (@ankur10)Photo by Ankur Jain (@ankur10)
Magical!
Real life, baby
1
59
Ankur Jain retweeted
Generally reason founders burn out - working on things you feel obligated to do but hate vs things you love
64
59
13
944
70,169
Ankur Jain retweeted
Look at that latency 😍 As proven by 25+ years of internet services, speed is a feature. For any app where there is a consumer (or agent!) on the other side waiting for the result, being even a bit faster results in higher conversion (and lower drop-off / churn) @wafer_ai is the choice for low latency inference 📈🚀
DeepSeek-V4.1-Flash is live on @OpenRouter!! pick @wafer_ai as your provider (ss taken 9.12.26)
15
14
1
61
14,364
For the first time ever, clean energy capacity in India has taken the top spot. That’s a 12x expansion over 25 years. Everyone’s talking about the environmental side of this shift, & that is big, of course, but I think the real story here is grid sovereignty. Scaling renewables at this pace moves energy away from being a volatile, imported commodity & turns it into a domestic advantage. When 1.4 billion people pivot their base energy capacity toward clean power, it completely redefines energy security. Huge milestone for India’s power landscape, & an even bigger blueprint for the rest of the world.
158
754
44
3,960
432,093
Ankur Jain retweeted
Who's gonna tell vibe coders about... > Rate Limiting > Caching > Load Balancing > Reverse Proxies > API Gateways > CI/CD > Docker > Kubernetes > Service Discovery > Circuit Breakers > Timeouts > Retries > Exponential Backoff > Idempotency > Message Queues > Pub/Sub > Event-Driven Architecture > Distributed Transactions > Saga Pattern > Dead Letter Queues > Cron Jobs > WebSockets > Long Polling > Server-Sent Events > Database Indexing > Query Optimization > N+1 Queries > Connection Pooling > Read Replicas > Sharding > Partitioning > Replication > Leader Election > CAP Theorem > Eventual Consistency > Optimistic Locking > Pessimistic Locking > Distributed Locks > Race Conditions > Deadlocks > Memory Leaks > Garbage Collection > Thread Safety > Backpressure > Autoscaling > Horizontal Scaling > Vertical Scaling > CDN > Edge Caching > Cache Invalidation > Feature Flags > Blue-Green Deployments > Canary Releases > Rolling Deployments > Rollbacks > Health Checks > Liveness & Readiness Probes > Monitoring > Logging > Distributed Tracing > Metrics > Alerting > SLOs > SLIs > Error Budgets > Observability > Secrets Management > IAM > OAuth > JWT Rotation > TLS > Encryption at Rest > Encryption in Transit > WAF > DDoS Protection > CORS > CSRF > SQL Injection > XSS > SSRF > Database Migrations > Schema Versioning > Disaster Recovery > Backups > Failover > Multi-Region Deployments > Chaos Engineering > Cost Optimization > Cold Starts > Serverless Limits > Latency > Throughput > P99 Latency > Tail Latency > Network Partitions > Clock Skew > DNS > TCP vs UDP > HTTP/2 & HTTP/3 > gRPC > Webhooks > API Versioning > Semantic Versioning > Infrastructure as Code > Terraform > Helm Charts > Build Caching > Dependency Hell > Production Incidents > On-call > Postmortems ...until the app gets its first million users.
935
672
239
6,770
774,501
Ankur Jain retweeted
the models have no moat (OpenAI, Anthropic, XAI) the IDEs have no moat (Cursor, Windsurf) the harnesses have no moat (Cognition, Factory, LangChain) the app builders have no moat (Replit, Lovable, Bolt) the wrappers have no moat (Harvey, Abridge, OpenEvidence) the inference providers have no moat (Together, Fireworks, Groq) the voice layer has no moat (Sierra, Decagon, ElevenLabs) the data labeling companies have no moat (Scale, Surge, Mercor) the AI infrastructure has no moat (Baseten, Modal, Railway) the neoclouds have no moat (CoreWeave, Lambda, Crusoe) the generative media companies have no moat (Runway, Higgsfield, Suno) apparently nobody in AI has a moat except the venture firm ☠️
362
252
154
3,626
580,649
Anthropic asked Claude to solve the Riemann hypothesis. A 167 year old math problem nobody has cracked. 🤯 It failed to solve it. But here is what happened. Claude tried 650 ideas. None worked. The person running it said keep going. Believe in yourself. That is it. That was the human input. Claude spun up 60 subagents. They ran 2,400 shell commands. Wrote hundreds of Python scripts. Checked each other's work. Downloaded 54 papers from arXiv to make sure nobody had already done what it found. Then it came back with something nobody expected. It improved a lower bound that mathematicians have been slowly pushing for decades. From 41.6% to 67.2%. A jump that big has never happened at once. Two mathematicians at Anthropic verified it. Two external experts checked it. Claude also wrote a formal proof that passes standard validation. And then it volunteered to write the paper. The part that genuinely matters here. The human who prompted this is not a mathematician. He just told Claude to try. And when Claude was skeptical, he said believe in yourself. That was enough. AI did not solve the biggest open problem in mathematics. But it just made more progress on a related problem in a day and a half than humans made in years. We are watching something happen in real time.
We asked an unreleased research version of Claude to take a stab at the Riemann hypothesis. It didn’t solve it, but it did make strides on a related problem: it increased the lower bound for the fraction of zeros of the Riemann zeta function that satisfy the hypothesis from 41.6% to 67.2%. anthropic.com/research/riema…
22
123
18
821
373,649
Ankur Jain retweeted
AI slop is everywhere. 69% of AI users admit to shipping work they haven’t verified, don’t fully understand, or couldn’t confidently defend, according to @glean's Work AI Index. And even when AI output is factually correct, it can still be generic, too long, or clearly not written by a human. The deeper problem is what researchers call “mental proof”: visible evidence that someone engaged with the work, cares about the result, and understands the person on the other end. When work feels like slop, it erodes trust and relationships. This is hard to solve. At Glean, we’re still figuring out how to keep AI from making our internal communication longer, blander, and less human. A few principles guide us: - 𝗜𝗻𝘃𝗲𝘀𝘁 𝗶𝗻 𝗮 𝗰𝗼𝗻𝘁𝗲𝘅𝘁-𝗿𝗶𝗰𝗵 𝗔𝗜 𝗽𝗹𝗮𝘁𝗳𝗼𝗿𝗺. AI can help protect against slop when it understands the context around the work. It should match the length of a communication to its complexity. When communication involves a sensitive relationship, ambiguity, or higher stakes, it should recognize that more human judgment may be needed and prompt the user to pause or add humanity. - 𝗟𝗲𝗮𝗱𝗲𝗿𝘀 𝘀𝗵𝗼𝘂𝗹𝗱 𝗺𝗼𝗱𝗲𝗹 𝘁𝗵𝗼𝘂𝗴𝗵𝘁𝗳𝘂𝗹 𝗔𝗜 𝘂𝘀𝗲, 𝗮𝗻𝗱 𝗺𝗮𝗸𝗲 𝗶𝘁 𝘀𝗮𝗳𝗲 𝘁𝗼 𝗽𝘂𝘀𝗵 𝗯𝗮𝗰𝗸. I try to model this myself, and push my team to do the same by saying something like: “This is too long. Please rewrite it more concisely.”  - 𝗧𝗵𝗲 𝗽𝗲𝗿𝘀𝗼𝗻 𝘄𝗵𝗼 𝘀𝗵𝗶𝗽𝘀 𝘁𝗵𝗲 𝗼𝘂𝘁𝗰𝗼𝗺𝗲 𝗺𝘂𝘀𝘁 𝗿𝗲𝗺𝗮𝗶𝗻 𝗮𝗰𝗰𝗼𝘂𝗻𝘁𝗮𝗯𝗹𝗲 𝗳𝗼𝗿 𝗶𝘁. Otherwise, it’s too easy to deflect responsibility to the AI. - 𝗠𝗲𝗮𝘀𝘂𝗿𝗲 𝘄𝗵𝗮𝘁 𝗺𝗮𝘁𝘁𝗲𝗿𝘀. If you measure vanity metrics like output volume instead of quality and outcomes, you’ll optimize for more output, even when that output creates more work for everyone else. More AI output does not create more value if people spend their time sorting through slop.
21
8
4
106
13,048
Congratulations @SkyrootA team! Proud of what you guys have done. India's first private company to reach the space.
We’ll be watching this ignition on loop all day. Kalam-1200 — Vikram-1's first stage, named for Dr. A.P.J. Abdul Kalam — roaring to life and lifting Vikram-1 off the pad, on a flight that reached orbit. #Vikram1 #MissionAagaman
1
122
Ankur Jain retweeted
.@pmarca on the "three-way Mexican standoff" between coders, PMs, and designers: "I'm seeing it in a bunch of the early leading edge companies in the Valley, they're circling around a job title loosely called 'builder,' or something like it. And basically the idea is that you had these separate jobs in the past of programmer, product manager, and designer." " I've been describing what's happening in the Valley companies as sort of this three-way Mexican standoff, where the programmers think that they don't need the product managers and the designers anymore 'cause they can have AI do that, and then each of the other two doesn't think they need the other two either."  "I've been predicting they're all correct. The product manager can generate code and design now, and each of them can do the job of all three. The idea is the jobs change so now the job is builder." @MTSlive
61
91
46
964
161,823
This is the part of the AI cycle where things start to get worrisome Capex commitments are unprecedented Funding of capital intensive business models is also unprecedented So much Capex will surely lead to use of leverage to boost RoE somewhere in the capital stack If a platform shift happens or technology leapfrogs (eg 10x better memory, or scale up of data centers in space, or even possibly non silicon chips? Yes some people are attempting these … 🤯) or even an over supply at some point .. that could make things complicated for those with big $ commitments For this reason, it feels crazy that investors have rotated to this extent out of software into hardware/capex centric businesses No question, even if there is some pain or correction … AI will be the biggest technology mega trend for our generation
7
7
1
136
15,377
The "AI job loss" narratives are all fake. AI = massive ramp in productivity = massive ramp in demand = massive jobs boom. Watch.
1. Employer recovery from post-COVID hiring correction 2. Employer recovery from post-COVID interest rate spike 3. Elasticity = demand boom QED
644
779
248
7,996
1,283,252
Ankur Jain retweeted
For decades, society has sold us a specific formula for success: High IQ + Perfect Test Scores + Specialized Knowledge = A Secure Future. We hire the “smartest” people. We obsess over grades. We treat intelligence as a scarce, gold-standard resource. But @JensenHuangcon is telling us that the era is over. I agree "taste" will be scarce. medium.com/activated-thinker…
96
165
37
984
207,807
An exciting milestone for AI in science: Our C2S-Scale 27B foundation model, built with @Yale and based on Gemma, generated a novel hypothesis about cancer cellular behavior, which scientists experimentally validated in living cells.  With more preclinical and clinical tests, this discovery may reveal a promising new pathway for developing therapies to fight cancer.
534
3,110
1,030
21,303
6,882,145
Graphene just broke a fundamental law of physics. Its electrons just did something physicists thought was impossible. For nearly 200 years, metals have obeyed the Wiedemann-Franz law – the rule that electrical conductivity and thermal conductivity always rise and fall together. But in ultra-clean graphene, researchers at the Indian Institute of Science found the opposite. As electrical conductivity increased, thermal conductivity dropped, shattering a principle taught in every physics textbook. The key lies at the “Dirac point,” a strange electronic tipping point where graphene is neither a metal nor an insulator. Here, electrons stop behaving like individual particles. Instead, they flow collectively as a nearly perfect fluid – a state called a “Dirac fluid.” This discovery doesn’t just rewrite the rules for graphene. It provides a tabletop window into extreme physics usually reserved for black holes and high-energy colliders. Scientists say this behavior could help probe mysteries of quantum entanglement, black hole thermodynamics, and the very fabric of matter itself. ["Universality in quantum critical flow of charge and heat in ultraclean graphene." Nature Physics, 2025]
274
1,519
174
7,157
542,439
We have to take the LLMs to school. When you open any textbook, you'll see three major types of information: 1. Background information / exposition. The meat of the textbook that explains concepts. As you attend over it, your brain is training on that data. This is equivalent to pretraining, where the model is reading the internet and accumulating background knowledge. 2. Worked problems with solutions. These are concrete examples of how an expert solves problems. They are demonstrations to be imitated. This is equivalent to supervised finetuning, where the model is finetuning on "ideal responses" for an Assistant, written by humans. 3. Practice problems. These are prompts to the student, usually without the solution, but always with the final answer. There are usually many, many of these at the end of each chapter. They are prompting the student to learn by trial & error - they have to try a bunch of stuff to get to the right answer. This is equivalent to reinforcement learning. We've subjected LLMs to a ton of 1 and 2, but 3 is a nascent, emerging frontier. When we're creating datasets for LLMs, it's no different from writing textbooks for them, with these 3 types of data. They have to read, and they have to practice.
378
1,727
203
11,744
723,451
I write in @FinancialXpress with @ankur10 @jiviai AI on emergence of #Artificialintelligence in #healthcare signifies one of the biggest shifts in modern #medicine. The article underscores need for #Accountability and himan touch to AI decision-making @IndiatechO
1
3
3
307
Ankur Jain retweeted
COMING UP AT 6 PM | Startup Central Don’t miss our interview with Ankur Jain from Jivi AI as Andrew Ng makes his first India investment in health-tech firm Jivi @sumitakareer @ankur10
1
3
6
2,530
Historic win for India as our chess contingent wins the 45th #FIDE Chess Olympiad! India has won the Gold in both open and women’s category at Chess Olympiad! Congratulations to our incredible Men's and Women's Chess Teams. This remarkable achievement marks a new chapter in India's sports trajectory. May this success inspire generations of chess enthusiasts to excel in the game.
730
12,222
240
83,459
3,416,612