HNHacker News
TopNewBestAskShowJobs

sashu123

38 karma · joined June 2, 2022

submissionscomments
sashu123··on U.S. on Track to Add $19T in New Debt over 10 Years
The simple answer is we are debasing our currency in a very complex way, that we cannot measure how much we debased it. In the past kings printed new coins to fund their wars or whatever and it was easier to track how much the currency lost value. In the present because economies have gotten complex + tools to debase have gotten complex we cannot estimate how much we have actually debased our currency by. If it’s 100% then there’s going to be a lot of inflationary pain, if it’s 10% we can probably handle it but I wouldn’t hold my breadth on getting any good estimate
sashu123··on Show HN: DocsGPT, open-source documentation assistant, fully aware of libraries
In this day and age there is no way a well funded University Lab can build a model that is even a fraction as good as GPT-3. Just look at "Bloom", from BigScience a commendable effort but useless for any real world use case compared to GPT-3. Unfortunately moving forward, if you want to build useful apps that use ML Models, you need to call API's from well funded Industry groups.
sashu123··on Big Tech was moving cautiously on AI. Then came ChatGPT
Blenderbot is so bad compared to ChatGPT. I find Yann Le Cun and facebooks take that they had something like ChatGPT but we’re just too cautious to release it to the world quite cringe. Why not just admit this upstart beat you, it will help you catch up if you diagnose the issue
sashu123··on NeevaAI
Perplexity.ai seems to give me better results for now
sashu123··on SpaceX said to fire employees involved in letter rebuking Elon Musk
Glad to finally see a CEO acting like a monarch and not a pathetic wimp. We need more such companies to restore order in the society
sashu123··on Sr Manager at Google Resigns After Dalit Activist Disallowed from Giving Lecture
> Most importantly, polling opinions about this is the worst method of determining if discrimination exists. A better method would look at objective measures like salary/job/hiring rate/layoff rate/etc of lower-caste Indians compared to Indians as a whole.

I don’t get this at all. If we treat a group as fairly as possible and on average they perform worse than other groups, is that evidence of discrimination? Is there a hypothesis that they need to perform equally as well.?

sashu123··on Google cancelled a talk on caste bias
fair skin thing is nonsense. I'm a Tamil Brahmin (Brahmin being the highest caste) and I'm one of the darkest skinned people in India. Most of my friends are far darker than most Indians from North India. It's much easier to identify through surname (Pichai I'm pretty sure is a brahmin surname), but certain states like Tamil Nadu banned surnames for this reason. If people ask me my surname, it is T. Every government document has T as my surname, I just go with that