KuriousCat
KuriousCat on Hacker News
They cannot reliably enforce copyright, so they fall back on terms of service and deplatforming.
Tried it with chatgpt, and of course it failed. It showed pulleys and belt rotating in opposite directions for a simple two pulley coupling.
If financial incentives align, we will be surprised how effective they would become.
I was asked by someone today if there are such startups with COBOL based code bases...
No, equating the number of publications to productivity in research is flawed and often frustrating, as it prioritizes quantity over deeper impact, quality, and genuine contributions. Higgs published fewer than 10 papers and he has a Nobel prize.
What does this even mean? I have to inform somebody that they can't use copyrighted material? Isn't it the whole point to begin with?
It’s baffling that on one hand we suspend in‑drive app interactions for safety, yet on the other, a product built to sidestep those safeguards and promote driver multitasking still attracts funding. I’m a little sorry to say it, but celebrating ‘move fast’ execution without deeper safety thinking is genuinely disturbing.
What is the model used by perplexity here?
You know what I would like even more, not having to pay tolls, parking etc having already paid for the tabs, registration and taxes…
It is usually worth the effort and would provide you with invaluable experiences…
I am not aware of any geographical constraints you might have but from what little I have seen, try picking up a good school either in Germany or Switzerland for your undergrad. Get a masters fellowship such as Erasmusmundus or DaaD, then move to Berkeley/Stanford for PhD…
As others have already mentioned, the post is a bit vague on details. However, what is your role? Are you the CEO or the CTO? Why is your cofounder able to steer the company? Higher stakes? Your silence? Extrovert/introvert issues? Did you happen to detail your cofounder why their chosen path is not good for the company? Do you have a structure in place to resolve cofounder conflicts? Do you still have your heart at that place? How do you know you are not mistaken/your cofounder is heading in a wrong direction?
Posted by
KuriousCat on Tuesday February 25, 2025 @04:34PM
from the self-post dept.
I an any business deal or negotiation, both parties negotiate and come to mutually agreeable terms. In case of software or digital purchases, it seems the providers can unilaterally impose terms, change them as they deem fit even when the user does not want to be bothered by the updated terms. It is even more pressing when it comes to AI models, the companies can scrape and use all the content...
Is there an implicit assumption that secretary suitability is fixed and the secretary or you do not change or evolve? What if having a few worse secretaries early on gets you in a much better shape? How are these dynamics accounted for?
It was more like the prep time, either get some savings or fork out some time. Time lines are not that deep to be honest. But if there are ways to improve it, please let me know. I would be glad to incorporate the suggestions/insights.
I am trying to put together a beginner friendly guide on startup funding, please let me know other materials/sources that I should be including...
Posted by
KuriousCat on Wednesday February 12, 2025 @02:11PM
from the anyonyashraya.com dept.
Posted by
KuriousCat on Sunday January 19, 2025 @02:36PM
from the self-post dept.
There is a strong tendency to integrate LLMs with existing apps. Barring few errors that could eventually get corrected, LLMs are fundamentally programmable computational environments. With that notion, every app could be an environment that is dynamically instantiated. Do we need to fundamentally change software delivery? Treat LLMs as dynamic computational environments, generate...
That’s very interesting but the question is how and why are we failing to grep the real value? Or if have allowed media to get away with anything and everything without a sliver of responsibility?
Posted by
KuriousCat on Saturday January 18, 2025 @04:20PM
from the self-post dept.
AI companies are currently training their models using all the data that they could find, much is being debated whether data uploaded into the private repositories remains untouched. For instance, are git private repositories currently being used? Or your data on cloud storage drives is being used? Or device users are assumed to opt-in and data in edge devices is pulled in. Keeping this trend...
Only small problem is that models are neither thinking nor understanding, I am not sure how this kind of wording is allowed with these models.
Isn’t it more of sign of their troubles? I would suspect their wrappers are falling apart slowly…
This is neat, but is there any drop in the performance of the Hailo or NVme? If my memory serves me correctly, there few few posts in hailo forums saying that they can’t guarantee the use of nvme hard disk with the module.
I think the idea of molecular computing has been around for a while, would we be seeing a scalable prototype this year? What are the current blockers?
Posted by
KuriousCat on Monday January 13, 2025 @05:41PM
from the jstor.org dept.
I am quite happy with endeavouros for the last couple of years, it even plays well with nvidia cards. Having played with multiple distros from the point of installing the Ubuntu from the cds they used to send, I definitely feel things have improved a lot. And a sincere thanks to all who’ve pushed things forward.
This is nice but is it not very well known? My cognitive science prof used to suggest similar experiments even before 2010.