sangnoir

39 ms·

661.
by sangnoir 2y ago
> result in a better experience for all parties involved Yu are mistaken in assuming businesses want better experiences for everyone, including workflows that may lose them revenue (like cancelations). The friction is intentional and par
662.
by sangnoir 2y ago
> Judges are still allowed to defer to experts Then why is congress be denied the same latitude (via agencies)?
663.
by sangnoir 2y ago
Why would you want an app for bluetooth headphones. To change settings and to update the firmware. There are no standards for these operations, so you'd have to resort to an an OEM specific app (stand-alone, or integrated into a the OS
664.
by sangnoir 2y ago
> I have negative desire to get a phone billing itself around AI and LLM features Google has been tastefully integrating AI into phones for a while now using opt-ins. The now-many-years old phone screening feature ensures the fear of spa
665.
by sangnoir 2y ago
That's even more niche: not only do people have to read beyond the headline, they must also be familiar with a nerdy TV show that first aired a decade ago ( Silicon Valley ).
666.
by sangnoir 2y ago
> we're complaining that companies use "our incredible journey" to actually reminisce about the journey Yes. For clear communication, the title, introduction, and conclusion should all align with the body of the message wi
667.
by sangnoir 2y ago
A side-project for a very famous early-stage startup VC fund that offers advice, networking and mentorship so that hackers aren't just winging it?
668.
by sangnoir 2y ago
Your favorite brand likely has Samsung parts in it, and you won't be asked to buy a Samsung car.
669.
by sangnoir 2y ago
Can you give examples of zombie institutions?
670.
by sangnoir 2y ago
> I don't think anyone has the moral authority to tell anyone else what they can take, especially terminal patients Yet euthanasia is a very fraught subject in the land of the free. The country may be coming less religious, but the
671.
by sangnoir 2y ago
> Why would they still be paying AWS rates? They are almost certainly not paying sticker prices. Above a certain size, companies tend to have bespoke prices and SLAs that are negotiated in confidence.
672.
by sangnoir 2y ago
I avoid working at organisations with NIH syndrome - if they are below a certain size (i.e. they lack a standing dev-eng team to support their homegrown K8s "equivalent"). Extra red flags if the said homegrown-system was developed
673.
by sangnoir 2y ago
Calibrations are far from impersonal. Managers absolutely play favorites in who they stick out their necks for.
674.
by sangnoir 2y ago
> feel like helping a coworker should be factored into your performance. This is how you get voting rings where there's an old guard on the team who help each other get good reviews, and a steady stream of sacrificial team members
675.
by sangnoir 2y ago
> I buy my computers to use them, not to be "efficienr" Not every product is meant for you. I bet folk who cool their computers with HVAC and draw power by the MW do care about efficiency a lot. Also, Zen 5 beats Intel in sin
676.
by sangnoir 2y ago
Go ahead and complete the thought in the context of the comment I was replying to and review if your "dunk" is conflicting with the point I'm making... Companies inconvenience and put their employees in danger (of varying lev
677.
by sangnoir 2y ago
> I never understood why the company would put its employees in danger... Like forcing them to drive to the office 2-5 days each week when they could continue working from home?
678.
by sangnoir 2y ago
> This is probably bad news for ChatGPT 5. I don't think it's that likely this co-founder would leave for a Anthropic if OpenAI were clearly in the lead. I'll play devil's advocate. People leave bad bosses all the ti
679.
by sangnoir 2y ago
It's hard to track a sliver of transparent 0.125mm optic fiber back to the operator, unless you're doing it Ariadne's-thread-style - but you'd have to find it first (post explosion).
680.
by sangnoir 2y ago
And yet we've had lots of drama in Linux and Redhat mailing lists, involving people paid to work on the respective projects, and using their work email.
681.
by sangnoir 2y ago
> because the point is to run it not on a Mac. Says who? My Mac Mini runs Linux as the host OS, this project allows me to run MacOS as a guest OS on Apple hardware on demand.
682.
by sangnoir 2y ago
> Apple's walled garden produces a lot of real benefits for its customers that are part of what make it successful, and dismantling their walled garden is going to harm consumers. What are those benefits, and why would they evaporat
683.
by sangnoir 2y ago
> A capital "A" is the same in any font; I can write a message in monospace and my interlocutor can read it in sans-serif, or tap in morse code and have it rendered in hand- writing, in all cases the difference is immaterial Ma
684.
by sangnoir 2y ago
Yes - that's ehat I meant, but mangled it it while editing.
685.
by sangnoir 2y ago
You'll probably find interesting threads and links at https://old.reddit.com/r/LocalLLaMa
686.
by sangnoir 2y ago
We've had a similar debate before, but the last time it about whether Linux device drivers based on non-public datasheets under NDA were actually open source. This debate occurred again over drivers that interact with binary blobs. I d
687.
by sangnoir 2y ago
There are a bunch of downstream fine-tuned and/or quantized models where people collaborate and share their recipes. In terms of contributing to Llama itself - I suspect Meta wants (or needs) code contributions at this time.
688.
by sangnoir 2y ago
> Cannot under any circumstances be circumvented. Tangentially - I find it fascinating how the originalists on the supreme court get fuzzy on what a "well regulated militia" meant at the time the amendment was written.
689.
by sangnoir 2y ago
> This isn't how it works. As the models are improved, they learn more about reality largely on their own. AI models aren't complete blackboxes to the people who develop them: there is careful thought behind the architecture, d
690.
by sangnoir 2y ago
> Better training, larger models, and larger datasets, will lead to models that Hypothetically, with enough information, one could predict the future (barring truly random events like radioactive decay). Generative AI is also constrained

More ›