m00x

2,261 karmajoined 10 tahun yang lalu

Director of robotics engineering @

comments

5 hari yang laludiscuss

Agentic coding is absolutely not a failure. It's just not the 10x that CEOs really wanted it to be.

We learned that some tasks don't really benefit from AI while others do. My team went from 7 people to 2 (went to new teams, no layoffs), and we're doing the same amount if not more work than we used to.

Is it more draining and lacking of focused work? Yes. Is it more money for the business? Yes.

12 hari yang laludiscuss

If you're calculating stock gains, taxes, etc. You'll want to do it on BigDecimal on line items, then sum it up. Cents could lose precision leading to a large difference on millions of line items.

If you're doing static payments or displaying a number, it doesn't matter if it's float amounts or integer cents since you're not doing any calculations anyway.

Stripe uses cents mostly because it's the most foolproof way that won't cause massive issues as a public API.

13 hari yang laludiscuss

Hyphens here are meant to be formatting. You would be correct if this was a literary piece, but handbooks and sheets don't need to use these rules.

13 hari yang laludiscuss

It's a good indicator but absolutely no guarantee. I've had to tell project managers this because they didn't understand this concept.

13 hari yang laludiscuss

BigDecimal should be used by almost everyone except for HFT since they're really slow.

13 hari yang laludiscuss

You can't do everything you need with an integer. There are values you might want to display or calculate with that are smaller than cents. In some places you'll need things like BigDecimal, which are immune to floating point errors in most cases.

It's also safe to return decimal values for displaying values.

bulan laludiscuss

It could be a much bigger MoE model

bulan laludiscuss

On modern hardware a cow page copy should only take 1-5ms. Redis forks to save the db to disk and it's been a solid design choice.

I guess it depends on how sensitive your application is to main thread pauses.

bulan laludiscuss

[flagged]

2 bulan yang laludiscuss

These LLM replies are really getting annoying.

2 bulan yang laludiscuss

Maybe it has to be fine-tuned per tool just like functiongemma

2 bulan yang laludiscuss

what are you talking about

2 bulan yang laludiscuss

Good artists get paid plenty. The hard part of art was always about either representing emotion, or a story. AI can't do that as it has no emotion, nor story.

2 bulan yang laludiscuss

> how you do one thing is how you do everything.

That simply isn't true though. It's not even possible to be true. Will a neurosurgeon put as much time in their cooking/cleaning/etc as they do their surgeries? There's not enough time/energy.

2 bulan yang laludiscuss

[flagged]

3 bulan yang laludiscuss

They are profitable to opex costs, but not capex costs with the current depreciation schedules, though those are now edging higher than expected.

3 bulan yang laludiscuss

That's GPT4 thinking. New models use tools to look at current events or latest versions, and rely very little on weight knowledge.

3 bulan yang laludiscuss

Mythos is basically this

3 bulan yang laludiscuss

It doesn't reveal that they have no clue of what they're talking about. They know that a good number of idiot CEOs will hear this and dictate their engineers spend more claude/codex money, which benefits them.

Jensen isn't an idiot, he's capitalizing on the market.

4 bulan yang laludiscuss

Minimax 2.7 is fine for most web stuff. It's slightly worse than Claude at backend, but works great for frontend.

They're all slop when the complexity is higher than a mid-tech intermediate engineer though.