mrg3_2013

Interesting to see how this is getting downvoted. Somewhat expected. Many more head would roll from this scandal. Bill Gates, Peter Thiel are just starters

meh..Just because he donated doesn't mean one should ignore or dilute the severity of alleged crimes. Infact, I would trade someone who doesn't commit any such acts and still does not donate over someone who donates but does worst of all the crimes.

That makes total sense now, indeed!

Had a chuckle at the mention of Stalin. Made me think. I would also think, the evils would be the one who would badly want to live forever, if an option was presented.

This resonates with me. Too much of anything loses value. This includes life. If there's no death, it would take special individuals to make sense out of it.

Not specific to this, but if you can solve issue with progressive lenses, you have a big market. Progressive lens sucks. They simply do not work and causes too much strain and practically unusable for daily operations. What would be great if there's a digital way where the lens changes depending on your activity. Driving, then go to long vision. In front of laptop - change to reading glass etc. I will be your customer if you solve this

This.

OpenAI continues to muddy the benchmarks, while Claude continues to improve their intelligence. Claude will win long term. It'd be wise to not rely on OpenAI at all. They are the first comers who will just burn cash and crash out I suspect.

Exactly. Citing cost has been an AWS play which worked during early days of cloud - so they are trying to stick to those plays. They don't work in AI world. No one would want a faster/cheap model that gives poor results (besides the cost of frontier model keeps coming down - so these are just dead initiative IMO).

On LLM, my experience with Claude has been much better than OpenAI models (though my use case is more on code generation)

Using LLMs via Bedrock is 10x more painful than using direct APIs. I could see cost consolidation via cloud marketplace a play - but I don't see Amazon's own LLM initiatives ever taking off. They should just lose those shops and buy one of the frontier models (while it is still cheap)

This is just Amazon's 'me too' play. Doubt anyone serious in LLM space would consider this

DOA

When marketing talks about price delta and not quality of the output, it is DOA. For LLMs, quality is a more important metric and Nova would always try to play catch with the leaderboard forever.

Looks interesting! I think cost of running resources would be more helpful

Interesting point! In my case, it is mostly from China. So I could do geo block

ok thanks.

Not sure what you meant. Leetcode type things solves a problem of past. Future will be copilots doing lot of grunt coding work and human value add would be instructing copilots the right way, bringing broader contextual information. Solving binary search tree problems during interviews will eventually go away (once old guards go away)

I continue to get puzzled with leetcode type things. Will coding exercises like this become obsolete someday or will the old guards continue to push for such things to be used during interviews? Today coding Copilots can easily replace leetcode "winners"

This is unreal. I have never seen anything this fast. How ? I mean, how can you physically ship the bits this fast, let alone a LLM.

Something about the UI. Doesn't work for me. May be I like openAI chat interface too much. Can someone bring their own data and train ? That would be crazy!

If paid, can't you use chatgpt ? I believe khanmigo is based on openai GPT4

Signs of things to come at a broader level in 2024.

Agree with you.

Exactly how made-up issues are created. Vocal minority fallacy

yeah, I felt the same

I have never heard or felt anyone talk about caste in my 25 years in US. I work with both academia as well as industry.

This is interesting! Congrats on launching! I would personally love to see just simple csv dump as input :) Would you be able to generate possible questions automatically ?

Kool-aid was mostly leaders ending regurgitating leadership principles and gaming it (without internalizing the greatness of it)

Underrated comment :)

It may be useful (perhaps even necessary) if you have to be on AWS. But it is a systemic problem however. Lot of these services have "need to launch something" sticker attached - so AWS keeps adding layers over layers and end up with something very few people understand. First comment is spot on (networking tends to be a cargo cult)! Certifications tend to be in architect's space (talkers) while the real coders (builders) who build systems hardly want to know these things.

P.S - FYI I am nX AWS certified myself, so part of the clan :)

lol. This really is the issue. AWS certifications are great way to know how services can be over engineered. Adding seemingly complex solutions to "prove" technical worthiness is an Amazonian thing

I love this post! Mainly because it resonates with me 100% and aligned with my thinking, as I go build my startup.