Ankur

[Ankur @AnkurCDas]

@Ktens October heat. Between end of monsoons and beginning of winter.

[Ankur @AnkurCDas]

@Ktens Gone through this last year. Don’t worry, but do get an annual check up done on the implant. You’ll get used to it in a week.

[Ankur @AnkurCDas]

@Ktens Oh it was delicious - just the right proportions

[Ankur @AnkurCDas]

Hot chocolate with Brandy and Kahlua! Perfect drink for a cold rainy day. “Elephant’s Treat” at The Elephant House Café, Edinburgh.
👌🏼

[Ankur @AnkurCDas]

@dhimantn Air India to Group A?

[David Sacks @DavidSacks]

Harvey is a great example of how American companies are building world-class specialized models: they took an open-source base (Kimi K3), post-trained it on legal data, and delivered state-of-the-art performance on legal benchmarks at a fraction of the cost of frontier models. Restrictions that kneecap open models would do nothing to stop Chinese labs from shipping the next Kimi. They would, however, cripple the ability of startups like Harvey to create high-performance, low-cost vertical models. Of course some of the closed labs would love this — it eliminates their competition.

[Brett Adcock @adcock_brett]

Today we're introducing Hark Handoff

Handoff has been independently verified as the best internet-use model ever built, outperforming ChatGPT 5.4 & Opus 4.8

While others focus on coding, we focus on everyday life: ordering food, booking flights, shopping, & navigating the web

[Ankur @AnkurCDas]

Couple of images from my Alpharetta mornings. The tall Loblolly Pines, native to the Piedmont region, were great company during my walks.

[Ankur @AnkurCDas]

@chamath #nologos

[Ankur @AnkurCDas]

@ARanganathan72 How come I remember that blue stamp so vividly! Ephemera of our times!

[Ankur @AnkurCDas]

AI disrupting fishing!! 🎣

[Ksenia Moskalenko @kseniam0s]

Founders Fund just backed a robot that kills fish.

Not software. @ShinkeiSystems built a physical robot for boats that uses AI to kill fish instantly, cutting stress and doubling shelf life.

The fish brand is sold at Erewhon and Michelin restaurants.

@foundersfund thesis: low AI bets, odd spaces, almost no competition.

Sometimes the strangest market is the most defensible one.

[Ankur @AnkurCDas]

https://t.co/5BwbcvypCz

[Ankur @AnkurCDas]

I can vouch for the flawless internet experience from Doha all the way to the US. It’s wonderful to be in touch family, friends and colleagues on long transatlantic flights. 👍🏼

[Ankur @AnkurCDas]

Claude Code Essentials for Beginners. 👍🏼👇🏼

[beamnxw ./ @beamnxw]

https://t.co/YGnnnBalU1

[Ankur @AnkurCDas]

@ranjan_vittal @khoslaventures Exactly what many mission-critical domains in the world needs. Wish you success in bringing this vision to life Ranjan 👍🏼

[Ankur @AnkurCDas]

“Free cooling” is not yet resolved for orbital data center scale. But on a long-enough time horizon, that and all of your projections could be achieved with Elon at the helm. All of this hype is based on one man’s audacious vision and proven execution skills. From an investor perspective though, there is an obvious ‘key man risk’.

[Ankur @AnkurCDas]

The constant “future speak” from tech CEOs is getting tiresome.

[Ankur @AnkurCDas]

https://t.co/Su72wjmvcN

[Ankur @AnkurCDas]

Local AI setups with open models and powerful consumer hardware is going to be a mainstay.

[starmex @starmexxx]

AMD CEO LISA SU HELD A MINI PC ON STAGE THAT RUNS A 235B MODEL AND REPLACES YOUR $440/MONTH AI STACK

amd's ryzen ai max+ 395 is the first x86 chip that runs a 200 billion parameter model on one piece of silicon. cpu and gpu share 128gb of unified memory, no separate graphics card needed

the gmktec evo-x2 runs qwen3 235b fully, deepseek v3 comfortably and llama 3.3 70b with headroom. on linux you get 110gb of usable vram out of 128gb

amd claimed the chip beat an nvidia rtx 5080 by more than 3x on deepseek r1 inference. a lunchbox sized pc outrunning a $1,000 discrete gpu on a real ai workload

a heavy ai user pays $200 for claude code max, $200 for chatgpt pro, $20 for cursor and $20 for gemini. that's $5,280 a year and the box pays itself off in 9 to 10 months

install ollama, pull the model, point claude code at localhost. same interface, nothing leaves the machine, nothing costs per request

bookmark this and read the article below

[Ankur @AnkurCDas]

@rgp_001 @TVMohandasPai Domain specific SLMs should be very much on the cards for India - that’s where applied AI use cases will flourish.