9,795 karma · joined January 10, 2020
Data Science, Rocks & Te Wai Pounamu, New Zealand.
Important: If you are looking at this profile because you thought something I posted was wrong, dismissive or disrespectful I would really like to hear from you so that I could understand how you got that impression. I will not get angry. It is important to me to communicate clearly and if something I said struck you that way then I failed and I would like to correct it. [Mostly borrowed from another profile].
Contact user name a t g mail
Unoptimised jit < optimised interpreter (at least in this instance)
They are working on it presumably because they think there will eventually be a speed ups in general or at least for certain popular workloads.
Of course there is zero chance New Zealand would ever challenge Australia or France for territory, and little chance we could ever claim economic rights.
Neither English nor Indian and I would rate Naan bread right at the top if I had a choice.
But I don’t think they are creating accounts to scrape paywalled data from the original source itself.
If you told me to eat rice instead of bread I would probably be just as horrified
Can you tell the difference between Japanese brands or only between Japanese and non Japanese origin rice?
Is rice a significant percentage of food cost in Japan? Coming from a culture that uses less rice, the cost of price verses the cost of protein makes it just about a rounding error when cooking.
Now butter on the other hand has just about become a luxury good in New Zealand when it was once cheap, plentiful and extremely high quality.
But will they keep releasing the weights or do an OpenAI and come up with a reason they can't release them anymore?
At the end of the day, even if they release the weights, they probably want to make money and leverage the brand by hosting the model API and the consumer mobile app.
I can't imagine demand would be greater for R2 than for R1 unless it was a major leap ahead. Maybe R2 is going to be a larger/less performant/more expensive model?
Deepseek could deploy in a US or EU datacenter ... but that would be admitting defeat.
Universities have this issue too, despite many offering students and staff Grammarly (Gen AI) while also trying to ban Gen AI.
It is curious that this is the case. If chargers were profitable a couple of years ago, you would expect more chargers and more profit today.
I haven’t noticed much growth in chargers where I live but I have noticed more EVs on the road.
> …
> I removed two lines. Where did that happen?
If you read the paper you will see they provide the original as well as the version missing information.
I did mention this in my comment too.
I am quite sure I could find your two missing lines if you provide me the full poem.
Given that you are a prolific commenter on HN, I am sure a LLM could be fine tuned to detect missing text from your comments without additional information. For example …
> WinForms is still around. There have been further tec lly just a big tire fire and about the best you can do is to ignore all of them and develop in WinForms.
It’s probably possible to detect missing information from “tec” until “lly”. But to know what is between is not possible for a human either, beyond plausible guesses.
In this benchmark they give the LLM the necessary information to determine what is missing. For example “here is a poem, here is a version of that same poem that may or may not be missing lines. Are any lines missing?
It’s more a tuning issue IMHO than an inherent weakness in LLMs.
If I was asked to find an omission in an ML paper, my brain compares it with other ML papers, it does not need to compare it to Star Ward, Top Gear, Greek history, pottery and the other 1000s of contexts I may know about.
There needs to be a moat. I see all these startups with a moat that is no bigger obstacle than a puddle.
A local (geographically to myself) pre-LLM (Convolutional Neural Network) company invested in a huge moat, building a large training set of human labeled data ... I knew the moment I saw an LLM that could ingest image that their moat had evaporated in an instant. A lot of the technical staff either left or lost their jobs but the company appears to be limping on hoping for an exit.
If a startup depends on a secret prompt and a few MCP servers, they have no moat.
I think the next wave of AI-based SaaS (rather than pure AI API providers) will be companies like the Legal IDE (Tritium [1]), recently featured on the homepage. Tools that innovate and use AI, but would still be innovate even without it.
[1] I have no affiliation whatsoever. I just liked the demo and the concept.
Anyone know if its a legitimate non-profit or one of the "sort of" non-profit?
From the press release [3] ...
> TTAM Research Institute (“TTAM”), a nonprofit public benefit corporation based in California ... The TTAM Research Institute is a nonprofit medical research organization dedicated to helping scientists and non-scientists join together to unravel the mysteries of DNA - the code of life. TTAM believes everyone should have the opportunity to access their individual genetic code and be empowered to contribute it to scientific research. The TTAM Research Institute was founded and is led by Anne Wojcicki.
[1] https://who.is/whois/ttamresearchinstitute.org
[2] https://web.archive.org/web/20250000000000*/https://ttamrese...
[3] https://investors.23andme.com/news-releases/news-release-det...
There could be many reasons he waited this long. Maybe he waited until he was retired and would not face blowback. Maybe he just has some free time.
It is very plausible that WebOS could have been an equal peer to iOS and Android. CEOs have killed off projects that might have been great commercial successes while perusing short term gains.
In a decade's time we might hear a story from inside ATI or AMD how they killed off their chance of beating CUDA for short term gains.
There is quite a few on Google Image search.
On the other hand they still seem to struggle!
It’s doable but a lot more effort. Microsoft did it with WSL1 and abandoned it with WSL2.
On the other hand I think it is pretty cool to have foxes on the roof.
Is this the first time Apple has offered something substantial for the App store fees beyond the SDK/Xcode and basic app distribution?
Is it a way to give developers a reason to limit distribution to only the official App Store, or will this be offered regardless of what store the app is downloaded from?
https://en.wikipedia.org/wiki/Project_Looking_Glass
Liquid Glass looks a lot like coming up with changes for the sake of them.
A better approach is to decide whether your municipality meets or exceeds guidelines (the 85% that do).
I trust my city (in New Zealand), but there are other cities I wouldn’t because their water infrastructure is old and under funded, or because of known problems in the recent past.
In my country we just have a name but you need an ID document to get anything done.
Do banks etc. consider an SSN number to be a password?
If I gave a classroom of under grad students a multiple choice test where no answers were correct, I can almost guarantee almost all the tests would be filled out.
Should GPT and other LLMs refuse to take a test?
In my experience it will answer with the closest answer, even if none of the options are even remotely correct.