cheshire_cat

Joined 12 years ago

Karma

730

Submissions

Member since

Jan 22

View on Hacker News

Recent submissions

Comment · 2 months ago

What problems did you encounter while using mise?

Comment · 2 months ago

You might be interested in these critiques of Huberman:

"So, Should You Trust Andrew Huberman?" by Dr. Andrea Love: https://slate.com/technology/2024/03/andrew-huberman-huberma...

"Is Self-Help Guru PhD a Womanizer and Pseudoscientific Nonsense Broker? | Andrew Huberman Analysis" by Dr. Todd Grande: https://www.youtube.com/watch?v=ANW3daJggoQ

He definitely has content that is helpful to people, but there's probably better sources for the same information.

Comment · 2 months ago

Would be interesting to see how Crunch [0][1] does on that benchmark. Any plans to add it to the benchmark?

[0] https://news.ycombinator.com/item?id=49252127

[1] https://wiki.call-cc.org/eggref/6/crunch

Comment · 4 months ago

What would a special Lumo agent bring to the table that you can't already get by using OpenCode with an European inference provider?

Comment · 4 months ago

By that logic links to the FT and the WaPo also couldn't be posted. Which I guess is a fair position to take, but seems too limiting to me.

Comment · 6 months ago

Wouldn't it be more efficient to convert the requirements these 200 markdown files into Playwright tests?

You could still use an LLM to write and extend the tests, but running the tests would be deterministic and would use less resources.

Comment · 6 months ago

Sounds promising, thanks for your report.

I didn't want to say that they're not cheaper to run, artificial analysis also shows that they're cheaper. My main point was about it being important to also look at token efficiency, not only cost per token, to get the full picture.

Comment · 6 months ago

While the cost are lower than frontier models there are two factors that make DS4 Pro and K2.6 not as cheap as they might look.

For DS4 Pro there's a discount going on for the official API, which sometimes gets overlooked and mixed up in discussions. Simon uses the full price in the comparison, so that's not an issue here.

The other issue is that DS4 Pro and K2.6 often use way more reasoning tokens than the frontier models. In my testing there are certain pathological cases where a request can cost the same as with a frontier model because they use so much more tokens. To be fair I'm using DS and kimi via 3rd party providers, so they might have issues with their setups.

But if you look at the Artificial Analysis pages of the models you'll see that DSv4 Pro uses 190M tokens and K2.6 170M tokens for their intelligence benchmark, while GPT 5.5 (high) only used 45M.[0][1][2]

I recommend looking at the "Intelligence vs. Cost to Run Artificial Analysis Intelligence Index" ("Intelligence vs Cost" in the UI). The open source models are still cheaper to run, but not by as much as you'd think just looking at the token prices.