binarymax

8 ms·

211.
by binarymax 1y ago
I think comparisons like these are fraught with issues, but are you sure? How much energy would you require to read and summarize 250k words?
212.
by binarymax 1y ago
Only works if you can also remove the browser.
213.
by binarymax 1y ago
I’ve thought about this, because I’ve been meaning to make the switch. The solution isn’t that hard, just keep the smartphone in your desk drawer and use it for OTPs and banking apps. It’s annoying to have two phones and two numbers, but
214.
by binarymax 1y ago
Thanks that makes sense. Strange why pop science articles don’t explain this critical point. I probably should have looked it up.
215.
by binarymax 1y ago
I have a total n00b question. Why would this be a “constant”? Wouldn’t different galaxies and different matter in the universe expand at different rates, and be an acceleration/deceleration, where one observation is the derivative or
216.
by binarymax 1y ago
Indeed! I'm just looking at JVector which I wasn't familiar with - looks cool. Have you tried it with the billion-scale competition? (not sure if that's still running)
217.
by binarymax 1y ago
https://ann-benchmarks.com is pretty good but I agree it needs an update. I'd like to see modern embedding dimensions (384, 768, 1536, etc.) as well as filters and combined read/write latencies.
218.
Petabyte-Class E2 SSDs Poised to Disrupt Warm Data Storage(storagereview.com)
14 points by binarymax 1y ago|5 comments
219.
by binarymax 1y ago
May I ask, which version or ORT are you using? Were the outputs identical to PyTorch outputs for the same image?
220.
by binarymax 1y ago
Looks awesome. I work closely with Multimodal search and have had trouble porting CLIP to ONNX and other formats due to the lack of multi-head attention operators. Are you using Python for the CLIP inference, or did you manage to port it
221.
by binarymax 1y ago
No we don’t. I’m American and I’ve never seen this format.
222.
by binarymax 1y ago
Really nice work. I appreciate the example with JSDOM as that’s exactly how I use readability, and this looks like a nice drop-in replacement. Question: How did you validate this? You say it works better than readability but I don’t see an
223.
by binarymax 1y ago
Anecdata but I talk to lots and lots of AI/ML/DS engineers. Everyone knows the current LLM architecture won't work for AGI. All the "reasoning" models are just pseudoreasoning and there is severe data leakage and
224.
Retirement: Bing Search APIs on August 11, 2025(azure.microsoft.com)
4 points by binarymax 1y ago|1 comments
225.
by binarymax 1y ago
The URL should be https://azure.microsoft.com/en-us/updates?id=492574 (@dang can you please update the post?) Microsoft is cancelling its Bing Search API, and the only option going forward will be "grounding"
226.
by binarymax 1y ago
When the options are "do what we tell you and get paid" vs getting laid off in the current climate, the choice isn't really a choice.
227.
by binarymax 1y ago
Thanks for the response… do you see a future where engineers are just prompting all the time? Do you see a timeline in which todays programming languages are “low level” and rarely coded by hand?
228.
by binarymax 1y ago
Absolutely the wrong take. We MUST think about what might happen in several years. Anyone who says we shouldn’t is not thinking about this technology correctly. I work on AI tech. I think about these things. If the teams at Microsoft o
229.
by binarymax 1y ago
So I need to ask: what is the overall goal of your project? What will you do in, say, 5 years from now?
230.
by binarymax 1y ago
Thanks for the advice! I wasn’t planning on using geoIP at all for several reasons - the main two being inaccuracies AND that I’d need to use a paid 3rd party.
231.
by binarymax 1y ago
Yes! I still want to default to accept-language, but asking is key. What is your pattern for giving the user an option? I like this article in that it rejects icons, but I don’t like how they arrive at one language for multiple regions ht
232.
by binarymax 1y ago
Timely. I’m working on this right now for a product and have everything in place using the accept-language header and then adding a url param. I’m currently agonizing about how to let the user change their language, because I want to respec
233.
by binarymax 1y ago
According to the article, Apple won the right to ban Epic, so what is the defeat you mentioned? To be clear I think Apple is in the wrong here and the App Store tax is absurd. But what court decision says that Epic won?
234.
by binarymax 2y ago
My friend’s startup is doing great stuff in this space to make a VNS device for consumers. https://zenbud.health/
235.
by binarymax 2y ago
I use it all the time. If it’s “good” depends more on your model for embeddings, but you do need to know a bit to tune the index. Whatever algo you choose, read the paper. If you’re using lucene HNSW, it will scale but will eat lots and l
236.
by binarymax 2y ago
It’s surprising how challenging this is for both Elastic and Opensearch. The problem is that it’s all configuration and no convention, so you need to roll everything yourself. There should be prescribed recipes to make this simpler. If you’
237.
by binarymax 2y ago
Why weaviate and not FAISS? The latter is faster and lighter.
238.
by binarymax 2y ago
O(n) is still OK for vector search if n isn't too large. Filesystem search solutions are currently terrible, with background indexing jobs and poor relevance. This won't scale for every file on your system but anything in your w
239.
by binarymax 2y ago
The best hackathons are community driven, like the annual Ludum Dare or the advent of code. If you're going to a highrise office building for a hackathon then yeah, it's going to be overly commercial.
240.
by binarymax 2y ago
Thanks for the info. That's unfortunate, but I can somewhat understand why FF takes that approach.

More ›