IanCal

25 ms·

541.
by IanCal 1y ago
I'm a bit confused, how is it redundant here? It's trying to make the best possible argument from one side that seems to be wrong. Instead of taking the argument at face value, it takes the most charitable understanding of it (not
542.
by IanCal 1y ago
There's a load of things I think are things many kids could do much earlier in maths, though my personal feeling may be off. My son seems to be taking very quickly to the feeling of maths so I'm trying to share some parts that I
543.
by IanCal 1y ago
Interesting, thanks, I'll have to explore that.
544.
by IanCal 1y ago
Images in general, nothing comes close to Gemini 2.5 for understanding scene composition. They perform segmentation and so you can even ask for things like masks of arbitrary things or bounding boxes.
545.
by IanCal 1y ago
As a counterpoint I asked that simple question to gpt5 in auto mode and it started replying in two seconds, wrote fast enough for me to scan the answer and gave me two solid links to read after. With thinking it took longer (just shy of two
546.
by IanCal 1y ago
Because the set of problems we make to be solvable with code is huge and the world is messy. Many of these things really are at a very high level of abstraction and the boiler plate feels boilerplatey but is actually slightly different in a
547.
by IanCal 1y ago
Is that vectors for edges or for searching the nodes? I’m talking about encoding the edges as vectors for traversal.
548.
by IanCal 1y ago
That doesn’t require developer time though. Also that time is needed regardless, do you think it’s the majority of time related to releasing a feature?
549.
by IanCal 1y ago
Managing and running some containers is really easy though. And running daemons? Don’t we all have loads of things running all the time? I find it easier to have the same interface for everything, where I can easily swap around ports.
550.
by IanCal 1y ago
I had the idea a few hours ago so I'm sure there are holes in this but my first idea is forming a graph where the relationship isn't a fixed label but a description that is then embedded as a vector. First of all, consider that in
551.
by IanCal 1y ago
I think it's a journey a lot of us have gone on, it's an appealing idea until you hit a variety of really annoying cases and where you are depends on how you end up trying to solve it. I'm maybe being unfair to the academic s
552.
by IanCal 1y ago
That's the one! It's not even that weird of a case compared to others but is an excellent example. Here's the history of the Paris example: https://en.wikipedia.org/wiki/University_of_Paris where there w
553.
by IanCal 1y ago
This seems to miss the other side of why all this failed before. Rdf has the same problems as the sql schemas with information scattered. What fields mean requires documentation. There - they have a name on a person. What name? Given? Legal
554.
by IanCal 1y ago
Even without tuning Claude is pretty solid at this, just give it the sparql endpoint as a tool call. Claude can generate this integration too.
555.
by IanCal 1y ago
Also you can get a lot more information from images than you think, and even more from video. Superresolution was the term iirc. You can’t make up information but you can use knowledge of the subject to accurately fill things in and other a
556.
by IanCal 1y ago
It's hard to overstate this. They perform segmentation and masking and provide information from that to the model and it helps enormously. Image understanding is still drastically lower than text performance, making glaring mistakes th
557.
by IanCal 1y ago
Is this not addressed by othellogpt?
558.
by IanCal 1y ago
If you’re talking about it being proven or disproven you’re misunderstanding the point of the thought experiment.
559.
by IanCal 1y ago
Tbh I find this view odd, and I wonder what people view as agi now. It used to be that we had extremely narrow pieces of AI and I remember being on a research project about architectures and just very basic “what’s going on?” was advanced.
560.
by IanCal 1y ago
If you don’t think this happens for simpler methods you’ve never deployed them. It’s the exact same problem on a classifier. Have you actually worked with these and are we discussing real world cases?
561.
by IanCal 1y ago
No? But repeated words can impact simple nlp setups. I’m not sure what case you’re concerned about where added text impacts classification with an LLM but added words shouldn’t with a different pipeline. > And NLP stands for natural lang
562.
by IanCal 1y ago
Happy to help - this is a thing I’ve employed multiple times for real cases. One big benefit is that it uses the cheapest and most understandable approaches for the majority of cases, and scales up quite nicely. It has a neat place for very
563.
by IanCal 1y ago
Thanks, I'd not realised that, which means I have no idea if the things I've done outside of the API are this new one or not. That does feel classic google.
564.
by IanCal 1y ago
It depends on a lot of things but to add to your possible setups you can potentially improve results by using simpler systems for first answers and falling back afterwards. For example: If contains cafe and not internet/cyber/etc
565.
by IanCal 1y ago
> But no amount of text appended to an input document, no matter how persuasive, can cause an NLP pipeline to change how it interprets the remainder of the document, Text added to a document can absolutely change how an NLP pipeline inte
566.
by IanCal 1y ago
This is where rates hide the issue. One in ten cases that take hours on a phone talking to a person with detailed background info and spoofed things is one issue. One in ten people that see a random message on social media is another. Like
567.
by IanCal 1y ago
Hmm could the old models generate images before? Had they hooked up imagen or something? I can make images on the Gemini site.
568.
by IanCal 1y ago
Google have been terrible at every single rollout I’ve ever seen them do. I see an announcement and it’s a waitlist. It says I can use it right now and I get a 404, or a waitlist, or it doesn’t work in my country. With the AI stuff more oft
569.
by IanCal 1y ago
Second order effects do not mean they're less important, perhaps that's the misunderstanding. > I couldn't care less about the technicalities cooked up by ofcom. Then you will be incapable of discussing it with anyone look
570.
by IanCal 1y ago
Again that is a second order thing and is also not true. If it was all examples by ofcom would include age verification. It’d be like saying an 18+ limit for buying booze means full DNA tracking because otherwise we don’t know if people a

More ›