meric_

4 ms·

1.
by meric_ 11d ago
https://typesafe.ai/blog/antibenchmaxxing But also effectively this is a classification model. It excels at specific certain types of workloads, and obviously will fail at others. Not really sure how one benchmarks thi
2.
by meric_ 12d ago
You're fooling yourself if the quest provides any advertising data value, let alone 600$ The device is simply just being sold for as a loss and is not a data harvesting operation. Some quick googling shows the ARPU of a meta user is ap
3.
by meric_ 13d ago
Eh I mean this is common default no? If you inspect element on many sites it links to the career page. Reddit, discord, etc. Maybe a little distasteful since I suppose since these WordPress sites are not owned by WordPress, but I wouldn&#x2
4.
by meric_ 15d ago
Their $ per mtok is quite high though for an indie project with no roi
5.
by meric_ 24d ago
Cline experiment: https://x.com/cline/status/2085237843379519737 Muse code: https://developer.meta.com/ai/resources/blog/build-with-muse... > Co-trained with the harness. Muse Co
6.
by meric_ 24d ago
You should use their harness. They trained it on multiple harnesses but have specifically optimized it for their harness. Cline also did an independent experiment w spark 1.2 where using the native harness makes it use fewer tokens / t
7.
by meric_ 1mo ago
I don't understand. Why is meta paying 70% while Google tiktok etc are paying 15-30% Are these not the same apps
8.
by meric_ 1mo ago
Because those trivial offences are a nuisance. BART spent millions implementing fare gates which so far, is at a huge loss, but it's dramatically lowered maintenance cost and crime on the transit. Similarly stores with loss prevention
9.
by meric_ 1mo ago
Im curious, how often is SSD / storage speed performance really useful? I feel like for many people it's akin to gigabyte wifi in that its nice to have, but not really particularly necessary
10.
by meric_ 1mo ago
OpenRouter is already the entrenched go-to router tbh. There's competitors (Vercel) but they're already well known. Stripe can also make them more competitive because they can waive / profit off the transaction cost. Consider
11.
by meric_ 2mo ago
IMO I disagree. I find the built in harness (database, browser use, etc.) quite good. The only downside (and a big one) is that by not being natively offered by OpenAI, Anthropic, etc. you're paying @ API billing and not plan billing,
12.
by meric_ 2mo ago
It's apache 2.0. Not exactly restrictive?
13.
by meric_ 2mo ago
A janitor doesn't need prediction markets. They could just short the stock (and profit way more that way)
14.
by meric_ 2mo ago
Do you know where you are lol? HN is ran by YC. All of the ads are YC ads for that reason
15.
by meric_ 2mo ago
I'm not sure why they even included Opus in the graph. This is effectively a flash/haiku model & should only be compared against Luna
16.
by meric_ 2mo ago
Lol is that a helmet? Clearly they're taking the Anthropic path and are safety pilling their models
17.
by meric_ 2mo ago
I'm wondering what people expect these companies to do? Human moderators? Sure, but Meta products have approximately half the worlds population using them. At this scale it's pretty mind bogglingly difficult to moderate down every
18.
by meric_ 2mo ago
Isn't this a non story? (At least from Flocks perspective) it seems mostly like the fault of the officers involved. The title makes it sound like they actively seeked to silence a journalist, but nothing in the actual article correlate
19.
by meric_ 2mo ago
Video is better. The models just chose older video models which are not SOTA
20.
by meric_ 3mo ago
It's the YC playbook. I guess it works, Corgi for example a "AI" insurance company with like only 5 real engineers and a bunch of growth people. Their main customer is other startups mostly YC. Same with Delve.
21.
by meric_ 3mo ago
Yes but my point is - Resource limits are a "recommendation" and are not strictly enforced - Significantly boosting resources up to 3 did not statistically shift performance results Sure for old tasks you could argue that now its
22.
by meric_ 3mo ago
No they didn't. Please read: https://www.anthropic.com/engineering/infrastructure-noise
23.
by meric_ 3mo ago
Huh? What are you talking about? https://www.anthropic.com/engineering/infrastructure-noise Is anthropic benchmark maxxing and cheating on terminal bench too? They don't follow the strict resource "limits&quo
24.
by meric_ 3mo ago
It seems to rank at around the same as nano banana (slightly higher) in blind A/B test benchmark but of course gpt image is a step above both right now
25.
by meric_ 3mo ago
They founded LeapMotion previously which was pretty big and totally unrelated to AI. They've been doing all sorts of shenanigans it seems
26.
by meric_ 4mo ago
I'd be very cautious how matching works. For some markets like sports it's trivial, but many politics or economics markets have minute rule differences that dramatically change what the actual market is betting on. Many markets ha
27.
by meric_ 4mo ago
It's not your money. Tax breaks are no ones money. No money is being sent for a tax break.
28.
by meric_ 5mo ago
None of those are standard and open except for Decent and UVI which does have third party developers albiet not many. Anecdotally the only dev I know on UVI is acoustic samples? Since Virharmonic moved off of UVI onto their own powered by g
29.
by meric_ 5mo ago
Kontakt has no competitors really. There are no standard and open competitors, the closest is maybe something like HISE but it's not even close from a marketshare perspective.
30.
by meric_ 5mo ago
Uhh have you seen the revenue growth of them? Oracle 22%, Amazon is 17% (if im reading it right) and Meta was 33% I don't think any of them are learning the lesson you think they are

More ›