pmehta94's profile picture. 🤖⚙️ | Previously @deloitte , @manningCICS

Pratik Mehta

@pmehta94

🤖⚙️ | Previously @deloitte , @manningCICS

Pratik Mehta reposted

Movies are great though. Even if you set aside the pure artistic enjoyment (you shouldn’t). Movies are stories, and stories are powerful, primal, moving, motivating. They are prompts to you to consider dilemmas and scenarios, to build your world model and compass. My rec is to go…


Gotta catch em all

Can't even argue

IAmAaronWill's tweet image. Can't even argue


Celebrity-watch headlines are crazy. 'Actor was spotted hiding tears at a funeral'. Really?


The mind is an excellent place to bend time and create shorter paths toward breakthroughs


LLMs are a force multiplier. You had better not be starting out with negative or low base stats ..


November rain 🌧️


PROLOG

Programming language you learned once but never touched again?



Pratik Mehta reposted

Why can’t programmers tell the difference between Halloween & Christmas? Because oct 31 = dec 25.


n > 3 labs sprouting up dedicated to improving access to RL environments


Pratik Mehta reposted

The first law of decision hygiene is that most decisions do not benefit from extra time.


Pratik Mehta reposted

Call it unhealthy if you want, but isolation is still the most undefeatable coping mechanism ever invented.


The problem formulation would be in terms of an MDP - not RL. RL is still a class of algorithms. Not sure why this is confusing.

PPO is one of the RL algorithms; not the magic algorithm for every RL problem same goes for GRPO / GSPO or any other current policy gradients algo rl as a problem formulation is excellent - agent sees state - agent takes action - agent gets reward and next state this is so…



✨🪔🎇😊


Pratik Mehta reposted

The only time you truly fail, is when you decide to give up.


That sweet sweet vindication


Something about not having the ability to just pick what you want to watch made the TV experience more memorable back in the day


Pratik Mehta reposted

💯 Tasks are not jobs.

"AI isn't replacing radiologists" good article Expectation: rapid progress in image recognition AI will delete radiology jobs (e.g. as famously predicted by Geoff Hinton now almost a decade ago). Reality: radiology is doing great and is growing. There are a lot of imo naive…



I suspect things will turn out this way in fields other than radiology as well

In 2016 Geoffrey Hinton said “we should stop training radiologists now" since AI would soon be better at their jobs. He was right: models have outperformed radiologists on benchmarks for ~a decade. Yet radiology jobs are at record highs, with an average salary of $520k. Why?

deenamousa's tweet image. In 2016 Geoffrey Hinton said “we should stop training radiologists now" since AI would soon be better at their jobs.

He was right: models have outperformed radiologists on benchmarks for ~a decade.

Yet radiology jobs are at record highs, with an average salary of $520k.

Why?


Pretty sure that's suggestive of supervised learning, not RL

An OpenAI executive expects the “entire economy” to become an “Reinforcement Learning Machine.” This implies that AI might train on recordings of how professionals in all fields handle day-to-day work on their devices. Details on this new era of AI training: • AI developers are…



Some of the experience reading great code that was previously only possible to gain by reading great code is now possible by letting Copilot LLMs fill it in a constrained space.


United States Trends

Loading...

Something went wrong.


Something went wrong.