NBJack

17 ms·

271.
by NBJack 2y ago
No, I'm not even interested in the surveillance here (though it is also a good comparison). Full disclosure, I'm very pessimistic on what current LLM models are capable of, and even more pessimistic on the impact of social network
272.
by NBJack 2y ago
Why settle for indifference to self-harm? https://www.cbsnews.com/news/google-ai-chatbot-threatening-m...
273.
by NBJack 2y ago
Ironically, this would lead to dramatically unhealthy social interactions and be ripe for abuse. You are literally creating a bubble of interactions with technologically enforced rose-colored glasses. Get the right person in charge of the e
274.
by NBJack 2y ago
It's GANs all the way down.
275.
by NBJack 2y ago
It's a sad state of bad features, unwanted behavior, reduced functionality, and regular regressions in behavior/performance. Honestly, it's an embarrassing situation for them that shows how much they've lost whatever way
276.
by NBJack 2y ago
> With some effort, you can use local account For now. Microsoft has been trying to close that for a while, and I suspect the only thing keeping some of it alive is the stark security requirements some of their corporate/business cu
277.
by NBJack 2y ago
Let's not forget regressions in gaming performance. This is my number one issue, and the devices I do have Windows 11 on arent changing my mind, and I suspect this holds true for many others out there. Why upgrade if I just get downgra
278.
by NBJack 2y ago
I'm wondering about the first example of device drivers. Whether we are talking about Linux or Windows, anything beyond generic device drivers (which push the problem to the hardware) will frequently be vendor specific. You often can&#
279.
by NBJack 2y ago
I think that's fair: the success stories are very rarely "yay, we did it!" but much more often "this single change was the sole focus of a team/multiple teams for X months, and the launch/release/fix was
280.
by NBJack 2y ago
> That would obviously inconvenience some people, but the gist of the criticism is that the US duplicates medical capacity for the sake of convenience. Or outright kill them due to a delay in a critical finding. This is more than a conve
281.
by NBJack 2y ago
This is how I learned about it!
282.
by NBJack 2y ago
Not to discourage you, but note it took a while before Google succumbed. Hopefully Kagi will hold out.
283.
by NBJack 2y ago
I've been using it for a while now. It is marginally better, but not exactly night and day. It seems to struggle with intent at times, and others I just get the same bland results as the free engines. The privacy is a big plus however.
284.
by NBJack 2y ago
I very much agree this is effectivity a 'honeymoon' period. Expect the SEO collective to shift focus on AI if the search approach becomes profitable in a few years. That said, given an "AI search" is estimated to be at l
285.
by NBJack 2y ago
What do you want to see in the playlist behavior? I admit I build them without much effort in YT Music and never felt constricted, but I may be missing out.
286.
by NBJack 2y ago
Especially compared to one of their core competitors in the US market: Amazon Music. I don't know who pissed off who in the world of Amazon, but I'm shocked at how broken the Amazon Music app has been for the last few years. Rando
287.
by NBJack 2y ago
Whoa. People remember this thing. I recall doing as much as I could with my Apple //c and a dot matrix printer. My art skills were crap, but that didn't stop me. Browsing the border styles, trying to use the custom.. somethin
288.
by NBJack 2y ago
I regret to inform you that waterfall planning is often considered a fail state in the toilet development world (and a messy one to clean up). I can however recommend the Spiral Model [1] as a lesser known Waterfall variation, which carri
289.
by NBJack 2y ago
Not a network security expert, but the level of traffic necessary to figure out whats real would probably trip other detection mechanisms in the process.
290.
by NBJack 2y ago
You've got multiple sources of pure oxygen on board, and potentially hundreds of passengers counting on a safe flight in a metal tube hurling through the sky thousands of feet off the ground at hundreds of miles an hour.
291.
by NBJack 2y ago
I have had several laptops over the years like this. Full shutdown and power on does not reset some problems, like missing audio, missing wifi, etc. For Lenovo devices, I have to go as far as using the 'recovery' button. This goes
292.
by NBJack 2y ago
I can't hold a LLM accountable for bad answers, nor can I (truly) correct them (in current models). Dont forget to take into account how damn expensive a single GPU/TPU actually is to purchase, install, and run for inference. And
293.
by NBJack 2y ago
Then we already have access to a cheaper, scalable, abundant, and (in most cases) renewable resource, at least compared to how much a few H100s cost. Take good care of them, and they'll probably outlast most a GPU's average lifesp
294.
by NBJack 2y ago
Hasn't this been posted before? It's cool, I agree, but it's been around for a few years. Has there been an update?
295.
by NBJack 2y ago
Consistent, algorithmic performance on basic tasks. A great example is the simple 'count how many letters' problem. If I prompt it with a word or phrase, and it gets it wrong, me pointing out the error should translate into a cons
296.
by NBJack 2y ago
No. But it won't stop the industry from trying. LLMs have no real sense of truth or hard evidence of logical thinking. Even the latest models still trip up on very basic tasks. I think they can be very entertaining, sure, but not pract
297.
by NBJack 2y ago
Prompting only works if the training corpus used for the LLM has critical mass on the concept. I agree that the model should have alternatives (i.e. preface the comment with the tag 'Nit:'), but bear in mind code review nit-pickin
298.
by NBJack 2y ago
"If you just let the game happen, it almost plays itself." The quote from the episode certainly makes me think about the "idle games" genre that has emerged in that last several years.
299.
by NBJack 2y ago
It doesn't surprise me; that sounds exactly like how C++ has progressed (sans the enterprise beans, hopefully). 10+ years ago I considered myself proficient at C++. It made sense, I didn't mind the pointers, and while the std libr
300.
by NBJack 2y ago
Let's not forget unless this is constantly uploading, hostile folks will be even more inclined to attempt assault to 'remove evidence' from the employee.

More ›