3abiton

7 ms·

121.
by 3abiton 9mo ago
Great list (totally agree with backup phone). I now want to know your 100+ gadgets
122.
by 3abiton 9mo ago
So now 2026 is the year of the linux console (I guess once proton adopts to the latest version of wine?)
123.
by 3abiton 9mo ago
I wonder if there is a way to integrate this woth home assistant for historical data collection, then use it with other sensors to build a prediction model or some usedul analysis. A bit less principled approach but voild be interesting!
124.
by 3abiton 9mo ago
Interesting to see reasoning models lacking behind their non-reasoning counterpart
125.
by 3abiton 9mo ago
And honestly in a lot of the cases bm25 has been the best approach used in many of the projects we deployed.
126.
by 3abiton 9mo ago
Yes, until google makes more changes, but this has been the eay for over a year now.
127.
by 3abiton 9mo ago
This is untrue in reality. Literally I used more than 5 banking apps, and few investement ones (including 1 in the US). I could log in to all of them through a browser, using a phone number 2FA, or a proprietary authenticator of the bank (a
128.
by 3abiton 9mo ago
I would not be surprised if Google is sponsoring a lot of this efffort targeting young devs, and "teaching about security". They basically positioning their services as "authenticators" of truth, despite it 100% being ca
129.
by 3abiton 9mo ago
Yes, the chicken and the egg problem. But here is the thing, the more adopters there are the more likely to get support. Not to mention the userbase will be mainly in the EU.
130.
by 3abiton 9mo ago
Not to mention, play integrity is being used a some sort of "anti cheats" by bank apps and other essential services. Even some government apps in the EU, essentially forcing you to be spied on by google. The worse part is that, yo
131.
by 3abiton 9mo ago
I see the confusion. My fault. Before the Mac M3, intel briefly used to have the intel core m3 cpu lineup. Fanless, and very energy efficient for the time (~2017).
132.
by 3abiton 9mo ago
> At this point, I am the old lady who is driving to a Target to buy gift cards and give them to Jared, who is the Amazon Customer Support specialist with a suspiciously heavy Indian accent, waiting on the phone. It happens to all of us.
133.
by 3abiton 9mo ago
> There is a saying about Gauss: when another mathematician came to show him a new result, Gauss would remark that he had already worked on it, open a drawer in his desk, and pull out a pile of papers on the same topic. As if phd student
134.
by 3abiton 9mo ago
> Anyone who uses an Android device that doesn't pass the attestation know this well. Speaking of which, tangentially, what are the latest workaround for play integrity these days?
135.
by 3abiton 9mo ago
> Why is Anthropic offering such favorable pricing to subscribers? I dunno. But they really want you to use the Claude Code™ CLI with that subscription, not the open-source OpenCode CLI. Because they are harvesting all the data they ca
136.
by 3abiton 9mo ago
Karpathy has a great intuitive style, but sometimes it's too dumbed down. If you come from adjacent fields, it might be a bit dragging, but it's always entertaining
137.
by 3abiton 10mo ago
That was fast. I was looking into recent HN datasets, and they are impossible find.
138.
by 3abiton 10mo ago
Yes, but Latner said multiple time it's closed until it matures (he apparently did this with llvm and swift too). So not unusal. His open source target is end of 2026. In all fairness, I have 0 doubts that he would deliver.
139.
by 3abiton 10mo ago
It's barely gaining adoption though. The lack of buzz is a chicken and egg issue for Mojo. I fiddled shortly with it (mainly to get it working some of my pythong scripts), and it was suprisingly easy. It'll shoot up one day for su
140.
by 3abiton 10mo ago
Honestly fail2ban is amazing. I might doa write up on the countless of attempts on my servers.
141.
by 3abiton 10mo ago
Is there an automated way of doing this?
142.
by 3abiton 10mo ago
Shout outto MIT open courses, and Stanford online for pioneerig this.
143.
by 3abiton 10mo ago
> I get the feeling that it was trained very differently from the other models It's actually based on a deepseek architecture just bigger size experts if I recall correctly.
144.
by 3abiton 10mo ago
What's the math on the $50k nvidia cluster? My understanding these things cost ~$8k and you can at least get 5 for $40k, that's around half a tb. That being said, for inference mac still remain the best, and the M5 Ultra will even
145.
by 3abiton 10mo ago
You could use llama.cpp rpc mode over "network" via usb4/thunderbolt connection
146.
by 3abiton 10mo ago
To be fair it was used a lot during my physics studies. I opted to use it afterwards for integrals and derivations, very powerful.
147.
by 3abiton 11mo ago
Tbh I think theseguidelines are just anticipating future trends.
148.
by 3abiton 11mo ago
> Comparing models, or even different versions of the same model, is a pseudo-scientific mess. Reminder that in most cases, it's impossible to know if there is cross-contamination from the test set of the public benchmarks, as most
149.
by 3abiton 11mo ago
Now that you mention that, I wonder if LSH would perform better with slightly higher memory footprint
150.
by 3abiton 1y ago
I read the abstract, while not familiar with the topic, how would we go about limiting the inpact?

More ›