rohansood15

4 ms·

31.
by rohansood15 5mo ago
Nope, HN changed the title. https://imgur.com/a/UgJqWEh
32.
by rohansood15 5mo ago
The title is Apple Silicon costs LESS than OpenRouter. Not sure why it got updated to this - maybe because I referenced the original HN post? Here's the full post: TLDR; When you consider batching, cache and input tokens, together with
33.
by rohansood15 5mo ago
That's some HN shenanigans, I swear I copy pasted my original title. https://imgur.com/a/UgJqWEh
34.
Apple Silicon costs less than OpenRouter(twitter.com)
64 points by rohansood15 5mo ago|22 comments
35.
by rohansood15 5mo ago
How do you define your productivity? Are you astronomically richer and/or freer now that you're so much more productive?
36.
by rohansood15 5mo ago
But you couldn't do it with the Claude app or the recently announced Claude Managed Agents I think.
37.
by rohansood15 5mo ago
Umm, why could this not just be Claude on AWS marketplace? Seems intentionally deceitful.
38.
by rohansood15 6mo ago
There are more software engineers outside the first-world than there are within.
39.
by rohansood15 6mo ago
You're right. They'd prefer to fire 7% of their team that did nothing wrong instead.
40.
by rohansood15 6mo ago
Nobody is stopping them from capping usage at 3x subscription price. Except themselves, because it'll ruin their revenue growth story once they stop selling dollars for cents.
41.
by rohansood15 6mo ago
TOS is not an impenetrable immunity shield.
42.
by rohansood15 6mo ago
I am engineer by trade. If I pushed an update which wrongly busted my customer's usage limits at a trillion dollar company, I would expect to get fired. Alongside my EM.
43.
by rohansood15 6mo ago
If Anthropic couldn't catch these issues before people started screaming at them, do we really believe 50% of software engineering jobs are going away?
44.
by rohansood15 6mo ago
Chinese companies are built on IP theft, and Anthropic/Open AI are not? And in world where code generation costs are trending to zero, goodluck commanding a premium to produce any kind of code. There is a whole bunch of P99 code that i
45.
by rohansood15 6mo ago
Most code is not P99 though. Also, have you considered that your trust in Anthropic and distrust in China may not be shared by many outside the US? There's a reason why Huawei is the largest supplier of 5G hardware globally.
46.
by rohansood15 6mo ago
You're assuming errors cannot be retried/recovered. They can.
47.
by rohansood15 6mo ago
Thanks for all the amazing work Daniel. I remember you guys being late to OH because you were working on weights released the night before - and it's great to see you guys keep up the speed!
48.
by rohansood15 6mo ago
Most open-source CVEs will have a patch linked in their disclosure. You can get vulnerable code via the git diff, then just verify if it is part of the LLM's finding.
49.
by rohansood15 6mo ago
I worked in AppSec in the past, made sense to me. Maybe you aren't the target audience? You don't really need manual verification for these, the CVEs (vulnerabilities) are public and can be programmatically validated.
50.
by rohansood15 6mo ago
Why do you think that making people do what they don't want to do for more money is the most effective way to distribute wealth?
51.
by rohansood15 7mo ago
This email gives out the endgame - eventually, Claude subscription would be ~30% cheaper than API costs. Our engineering team averages 1.5k per dev per month on credit costs, without busting Max limits today.
52.
by rohansood15 7mo ago
The paper is about vector quantization, which affects KV cache not model weights/sizes.
53.
by rohansood15 7mo ago
Thank you.
54.
Private LLM Inference on Consumer Blackwell GPUs(arxiv.org)
3 points by rohansood15 7mo ago|0 comments
55.
by rohansood15 8mo ago
What's the cheaper alternative from Gemini for Flash-2.5-lite level intelligence when it gets deprecated on 22nd July 2026?
56.
by rohansood15 8mo ago
Yea but there is a whole world of tasks for which Flash 2.5-lite was sufficiently intelligent. Given Google's depreciation policy, there will soon be no way to get that intelligence at that price.
57.
by rohansood15 8mo ago
For the last 2 years, startup wisdom has been that models will continue to get cheaper and better. Claude first, and now Gemini has shown that it's not the case. We priced an enterprise contract using Flash 1.5 pricing last summer, and
58.
by rohansood15 8mo ago
+1 to this. Been using Codex the last few months, and this morning I asked it to plan a change. It gave me generic instructions like 'Check if you're using X' or 'Determine if logic is doing Y' - I was like WTF.
59.
by rohansood15 9mo ago
Playwright codegen.
60.
by rohansood15 9mo ago
> Would you trust a Cursor review of Claude-written code more, less, or the same as a Cursor review of Cursor-written code? You're assuming models/prompts insist on a previous iteration of their work being right. They don'

More ›