Angik Sarkar
@kyunbit
Followers
1K
Following
17K
Media
141
Statuses
5K
Co-founder & CEO, @ixanaHQ. Developing high-speed HCI Past: Founder, Waylo (acq @edreams). @intel Interested in physics, AR/VR, machine learning, SoC, HCI
Seattle
Joined April 2009
How you interact with computers will completely change in the next decade! Watch the video with the sound on. Introducing Wi-R: wire-like wireless that communicates with touch. Science fiction is now a reality!
8
7
43
Chatgpt Image seems incredibly good at converting mermaid schema to good looking flowchart. Grok almost always messes up the text. Gemini flowcharts look bland with occasional text mistakes.
0
0
0
Wi-R NFE (Near Field Electric) changes how you interact with devices around you. Read more here: https://t.co/lmrqySrbFM
0
0
2
Apart from tab completions, I really like Antigravity. Am I the only one?
0
0
0
Loved the energy at SPC India.
@kyunbit from @IxanaHQ showed us how they're using the human body as a wire to stream high speed data! https://t.co/EfOr7z1wbS
0
0
4
One of my pet peeves about Hollywood: smart people are depicted as having 3 PhDs who know everything about every topic and never fail to solve a problem in an hour. If we want to become a tech forward superpower, we need to change public perception about how science/tech is
Anduril’s autonomous weapons stumble in tests and combat, WSJ reports
0
0
1
Wow. This is exactly why I learned not to leave any notes, unless it's a removable paper stuck between pages
Recently a friend observed that I don’t make any notes on the books I read and asked me why. I couldn’t answer promptly but remembered this was hard wiring from school days. Resale value of textbooks would go down materially if you had made a mess of it. It’s fascinating how
0
0
0
Finally finished watching Dropout. SV's common refrain is that few SV investors backed Holmes. Can't help feeling that the world might have benefitted from @ElizabethHolmes' passion if she had Tier 1 VC and mentors. With helpful and patient investors, the science risks and the
1
0
1
A great nuanced take. Every baby is a superbaby, some just in skills that are more worshipped by the external world.
I'm going to weigh in on Nucleus. It's always tough to see a pile-on of this magnitude. Nucleus has a partnership with Illumina so you can get deep sequenced for $500. That's pretty cool, and why I bought a kit a while ago. I want more people to do wgs (whole genome sequencing)
0
0
2
The intersection of “seems like a bad idea” and “is actually a good idea” is the sweet spot for startups. "If a good idea were obviously good, someone else would already have done it. So the most successful founders tend to work on ideas that few beside them realize are good."
30
42
360
I wish more founders understood this. 1-2M raise is a suicide raise. If you really want to raise that, start low e.g. $750k Angel round. Let the round oversubscribe to $1.5M.
understanding fund dynamics for founders is almost as important as your customers every fund has a box they need to check: - $250M fund: minimum $5M check, needs 15% ownership - $25M micro-fund: maximum $500k check, want 7-10% - $1B growth fund: minimum $15M, need board seat
0
0
4
you know why Valar Atomics is going to win? it’s not just that we built WardZero in 10 months. it’s not just because we’re the most capital efficient nuclear company in the world. it’s not just because we split the atom. It is all of these things, but here’s what struck me the
Today, Valar Atomics became the first startup in history to split the atom. Announcing Project Nova, a series of zero power critical tests on Valar Atomics' Nova Core in collaboration with Los Alamos NCERC and NNSS. Nova went critical for the first time this morning at 11:45am.
26
48
916
Grit isn’t showing up in the office 9-9-6 when the VC dollars are flowing and ACV doubles every week; it’s pushing through and showing up when everything is falling apart but you still believe
16
27
326
Wouldn't it be better if the mortgages were portable instead? 30 to 50 year change will change mortgage payment by a few hundred/month for most people. Portable mortgages would open up liquidity in the market. Demand and supply then decide the prices.
Thanks to President Trump, we are indeed working on The 50 year Mortgage - a complete game changer.
0
0
0
startups -- in most cases, a really high win rate > 50% is actually NOT a good thing ... 1. price too low 2. not a healthy mix of outbound 3. a signal you're in a small market
16
6
113
Conditional sampling and denoising via thermal relaxation with constraints. Probably something similar to what the brain does for our perception.
Experiment: Thermodynamic Spin Inpainting We hide the middle of a letter "A" and let an (emulated) thermodynamic computer nudge the spins until the gap heals. The physics-y sampler literally “re-grows” the missing strokes. Under the hood: I used THRML to run block-Gibbs
7
14
120