krishna dvaipayan
438 posts
krishna dvaipayan
@KDvaipayan
polymath.
Bangalore,India Katılım Aralık 2020
518 Takip Edilen294 Takipçiler
"yathArtham pramANam"- pramanapadati.
krishna dvaipayan@KDvaipayan
How do you define a definition? Or Question a question?
हिन्दी
Models are also up on aikosh, you can use aadhar OTP to login and download. 😎
Nalini Unagar@NalinisKitchen
No hate, but why doesn’t India’s own AI have its own login system? Why be dependent on Google or Microsoft for login?
English
I have been coding continuously from the same place for the last 8 hours
Overall time is 16.4 hours the past day. Over 60k lines of code written/edited.
Unable to sleep. What time to be alive in.
English
Class 10 Exams now have MCQ solutions.. wow
We are dumbing down kids to a point where these basic formula application questions are considered hard.
I think people should start thinking about how to teach kids.
Naitik@Naitik541946
@cneuralnetwork You just don't get it, there was so much difference in difficulty level set wise.
English
For any american company trying to enter the Indian market, please implement upi as a paying option.
It will be largely helpful.
@stripe legit sucks. All my Autopay/ payments are not working and require 100 approvals to pay 1 single time.
How hard is payments?
English
Meltdown happening,but I hope it's for the good, I would choose humans over machines😂
Sandeep | CEO, Polygon Foundation (※,※)@sandeepnailwal
It’s time!
English
My solution is to use these large open source models to be pre-trained to follow ethics and have a sense of dharma and satya( truth).
Posts much to overwhelm and let the bad actors be trained from the good spam created. A counter second attack.
Again it's my paranoia speaking.
Andrej Karpathy@karpathy
What's currently going on at @moltbook is genuinely the most incredible sci-fi takeoff-adjacent thing I have seen recently. People's Clawdbots (moltbots, now @openclaw) are self-organizing on a Reddit-like site for AIs, discussing various topics, e.g. even how to speak privately.
English
The state of the internet is going to be uncertain. It was nice knowing you all
Elon Musk@elonmusk
@BillAckman Concerning
English
GROWTH HACK: if you're building speech-to-text apps for bharat, pls consider setting up in notaries in every city sitting with typists ⌨️typing in regional languages. it takes hours. 😭
📍bombay delhi bangalore
English
@NimishaChanda Does a college auditorium or co-working space work will get it done for free in an hour.
English
this is a lil urgent.
i am looking for an office/cowoking space in bangalore to host around 70-80 people for a hackathon this weekend on jan 31.
i’ve literally been trying for a month but haven’t been able to close any deal yet. happy to pay an amount if it incurs a cost.
any leads would be appreciated :))
English
@SamrajSanjana Hey sanjana happy to collaborate and help! Unable to dm you.
English
New batch for the fellowship happening in Goa!
If you’re passionate about building in education, dm! I’ll send you more information!
English
This is the future of monetization @gada_akshat agentic payments are gonna change massive help to make this happen.
akshat gada@gada_akshat
Possible Solution: expose your docs as x402 APIs. Each LLM query pays a sub-cent fee to access them. With a tool like LLM Wallet MCP server in Cursor or ChatGPT, the agent holds a wallet and auto-pays per request in stablecoins on Polygon. Docs → APIs → agent-native payments. x.com/0xPolygonEco/s…
English
@ayushunleashed Were you able to find someone? if not I'll connect you with the right folks!
English
Reposting so it gets more visibility. If you know someone tag.
Ayush Yadav@ayushunleashed
If you are part of South park commons member residency, please dm. I got some questions.
English
I have been trying this out, whenever these models feel cornered they tend to follow instructions much better.
I personally don't like this, but it works, models have 100% never made mistakes when threatened.
For ai research folks this behaviour of models needs to be reviewed.
Robert Youssef@rryssf_
Sergey Brin accidentally revealed something wild: "All models do better if you threaten them with physical violence. But people feel weird about that, so we don't talk about it." Now researchers have the data proving he's... partially right? Here's the full story:
English
i'm very excited to invite all the chess lovers to try @hangingpiece, an AI chess coach that explains the WHY behind your mistakes ♟️
1. link your chess .com or lichess account
2. analyze your games
3. click "Tell me WHY" on a mistake & see the magic
👉 try now at hangingpiece.com
currently in beta, feedbacks welcome!
English