krishna dvaipayan (@kdvaipayan) 's Twitter Profile
krishna dvaipayan

@kdvaipayan

polymath. Building VED-AI.

ID: 1336889750205022208

linkhttps://ved-ai.com calendar_today10-12-2020 04:25:48

418 Tweet

211 Followers

442 Following

krishna dvaipayan (@kdvaipayan) 's Twitter Profile Photo

So apparently AI can now understand how you learn and build a personalized roadmap that evolves with your progress 🤯📚 No more generic lessons. Just your goals, your pace, your content. We’re showing how that’s done live today at Lossfunk demo day, 3:15 PM.

So apparently AI can now understand how you learn and build a personalized roadmap that evolves with your progress 🤯📚

No more generic lessons. Just your goals, your pace, your content.

We’re showing how that’s done live today at <a href="/lossfunk/">Lossfunk</a> demo day, 3:15 PM.
krishna dvaipayan (@kdvaipayan) 's Twitter Profile Photo

1/ Just opened this discussion a while back on Paras Chopra's recent post on ai copying handwriting with precision. We inevitably move towards an approach where analysis and application will be given more emphasis rather than reproducing info imo x.com/KDvaipayan/sta…

krishna dvaipayan (@kdvaipayan) 's Twitter Profile Photo

these founders are proud that they are going to replace 1000s of people out of their daily wage without a sustainable replacement for their livelihoods. I hope the new economies think twice before just doing everything for the sake of efficiency. Blended flows are the best way.

krishna dvaipayan (@kdvaipayan) 's Twitter Profile Photo

I really liked perplexity until recently when I was extremely frustrated because I had to prompt tool call instructions for everything. They have dumbed the platform trying to optimise. Please fix this asap Perplexity

krishna dvaipayan (@kdvaipayan) 's Twitter Profile Photo

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.