Intern at PTC
Joined May 2021
- Tweets 262
- Following 54
- Followers 33
- Likes 50
24 Photos and videos
Devyani Chavan retweeted
May 26
🚨 Anthropic just showed a 27-minute workshop on how to actually do prompts for Claude.
Taught by the people who built it.
Free. No registration. No paywall.
I've seen $300 courses that don't cover what they teach in the first 8 minutes.
Watch it and bookmark it now.
34
239
1,341
289,343
May 26
don't have anything to say, just checking off my to-do list!! night!
1
1
48
Mar 22
hardly solved the first one, it seemed logical and easy but when it came to putting that logic down in code..... aghhhhhhhhhhh
82
Mar 20
people will spend days trying to squeeze 0.1% more performance out of a model, meanwhile a quick data quality check or a thoughtful feature could give 5%. it's the 90/10 rule of ML: impact mostly lives in the grunt work, not the fancy bits.
1
62
Mar 16
so many 'new' llms are just deepseek v3 derivatives or similar. feels like the core architecture problem is mostly solved, and we're just fiddling with normalization for stability.
sebastianraschka.com/llm-arc…
65
Mar 15
everyone wants a perfect model, but getting it into prod with decent latency usually means ditching half the fancy stuff.
34
Mar 15
so many general engineering roles getting cut, while companies hype AI automation.
32
Mar 14
academically, ml students are sharp on models. practically, watching them debug a broken data pipeline or provision infra for their 'simple' model makes it obvious the actual systems building gap is still huge. kinda wild.
57
Mar 13
this constant push for the absolute bleeding edge model is exhausting. most of us are still just trying to get basic, *reliable* inference at scale without the whole thing falling over. incremental stability over speculative leaps, every time.
72
Mar 13
meta ai drafts marketplace replies. the first 'intelligence' we push is always for transactional grunt work. what's left to teach?
85
Mar 13
product teams claim 'data-driven' until the data contradicts their pet feature idea. then it's 'qualitative insights' all of a sudden. never fails.
74
Mar 13
weird paradox: the higher your perceived status, the less original stuff you ship. reputation management quickly becomes the critical path, not actual building. that's a broken system.
Source: sharif.io/looking-stupid
68
Mar 12
can't get good sensor data for flash floods, so google had an llm read 5 million news articles to make its own training set. that's one hell of a data-gap workaround.
Source: techcrunch.com/2026/03/12/go…
1
50
Mar 12
starting to think the main feature of X is not tweeting… it's roasting people in the replies
2
50