@iruletheworldmo I don't know what God blessed team Anthropic seems to have assembled, but I've been plain shocked at how well Claude keeps improving when compared to other frontier models. It seems like the rest can't ever dare to catch up, because the foundation Anthropic has laid is beastly.
On days that I feel a little depressed reading bad news etc, the world and India feels hopeful back again due to gentlemen like Professor Kamakoti's presence. I have had the privilege of having had multiple interactions with Bala Periyava when I was young. This video made my day!
I spent a significant amount of time in the last 6 weeks building systems that help understand how money is moving in the Options world in the US stock market with the help of Claude. I was resisting doing this but eventually had to give in due to a size-able footprint I maintain in the US stock market.
This further dented the momentum in my primary field of interest, which is Robotics but the digression felt worth it, with my margin and IRA accounts now connected via Python scripts for trade automation.
Stock market engagement, especially the all mighty US one, can be very rewarding but an extraordinary amount of knowledge and maturity is needed to be able to consistently make profits without losing money. Options is easily one of the most prolific and dangerous instruments to deal with, so one has to tread carefully. My Python scripts are aggregating data from about 5 different data sources, to be able to provide daily signals.
With a good chunk of Python automation out of the way, slowly creeping back to Robotics again now, building and reading up https://t.co/tyC3vi4vso a great way to deepen the foundational knowledge.
With newer and newer data being generated by humans everyday, ingesting all that data and doing complex reasoning with it, might not become cheap enough for daily usage despite the advancements thus far. If it does become cheap enough, I think more complex forms of higher order work, that are even more ambitious, will be attempted across the world.
@narayananh Feels nice to see the photo you have posted. If there's no congregation happening here close by, I think I will start one up myself starting next year.
@rvaidya2000 It's such an extraordinary coincidence that I came to know about Shri @crkesavan as being a great grandson of Shri Rajaji on the very day that this article was published. Heartening to know that great legacies continue to live in one form or the other.
Though it may seem a long and bumpy road at the moment, with enough resources, I can see that this endeavor could lead to innovation that shifts the ground underneath massively, with ramifications in health, personal security etc. Kudos to the extraordinary decision, see you in 5 years with breakthroughs!
Beyond the idea of generating and retaining "jobs", there should be a system for rewarding good karma during times when humans are not able to find "jobs".
Every UHNI that wants a healthy society to exist for their next generation sake or out of goodwill, pools money to create a corpus from which monetary rewards are made for time spent by a needy human for a cause of their choice, say for physical help, religious participation, healthy lifestyle, learning a new skill etc.
Sort of like UBI for good deeds only. Full automation must arrive and people should be empowered to pursue their interests rather than hold onto boring and mundane "jobs". Weird that one has to hold onto a "job" to survive in life.
I never get a reply from you @svembu but it doesn't matter. LLM's seem capable of handing text well enough to be able catch graft within government related systems. If every digital request lifecycle submitted to a govt department comes under the scrutiny of, say a fine tuned Sarvam model, there's little chance for malpractice to go through. Can Zoho work with Sarvam in this direction if it's not already started? @pratykumar
Narrowing down AI's vast set of potential applications to just code generation is mere short term thinking prevalent today. There's a lot more use-cases including health monitoring etc that will start showing up. Indeed more efficient and reliable AI systems are possible, but the demand surge will be so high that it will outdo the currently planned infrastructure even if that happens.
Having said this, given the resource constraints in India, I agree that it makes sense to wait and watch how this plays out.