torginus

20 ms·

511.
by torginus 6mo ago
Aren't they 'wasting' time on these features exactly because the engineering requires a different, more traditional skillset from the ML work model people do, and can be done in parallel?
512.
by torginus 6mo ago
> Developing nations have the strongest incentive to cheat since they need those margins to catch up This isn't really how it works, since greenhouse gas output is pretty much corellated to income level, and even that's an unde
513.
by torginus 6mo ago
I did try this for syncing stuff between PCs, and have found the performance quite poor, like 10MB/s on a wired 10Gbe network. I don't know what the reason for this, but I suspect its networking layer is not really optimized. In c
514.
by torginus 6mo ago
It's not artist replacement yet because they dont have the necessary training or sophistication. I doubt the current state shows the end of their ambitions.
515.
by torginus 6mo ago
I think your monitor or room lighting might just be different from others'
516.
by torginus 6mo ago
I don't want to judge the US armed forces, so the following is a hypothetical - I heard the rumor that some procurement guy shared the US Army paid about $10k for a set of power tools for mechanics, which were commercially available, a
517.
by torginus 6mo ago
Yeah considering we're somewhat far along the maturity curve with LLMs, and diffusion, we can kind of extrapolate where this is going, so unless there's another gamechanging breakthrough, I can't connect the dots between what
518.
by torginus 6mo ago
I don't think the MIC is that lucrative even in the US, compared to the private sector. Consider the F35 - there have been like 1500 planes, and at $100mm a piece, the total comes out to $150B over two decades, and that's revenue
519.
by torginus 6mo ago
I'm not sure how intelligence and killing go together. The soldiers who are getting blown up by those suicide drones with just enough intelligence to recognize targets and chase them, are infinitely more intelligent than said drones. I
520.
by torginus 6mo ago
This feels like what a dog does. It's incredibly hard to train dogs by punishment, because it's very hard to tell if the dog understands what he did wrong and feels genuine remorse, or is just showing submissive signs at your disp
521.
by torginus 6mo ago
What? I thought Azure will always have the Sharepoint/Office/Active Directory cash cow.
522.
by torginus 6mo ago
I wish there was more research (maybe philosophy) would go into characterizing consciousness and intelligence, so that we could at least define what we are missing in current AI systems.
523.
by torginus 6mo ago
Btw, this makes a great argument for workers' rights - if you have a company who owns datacenters - well, you can't fire your GPUs to make your Q2 look better
524.
by torginus 6mo ago
I feel like there's a gross mischaracterization here - $14B in AI terms is like 'second-best VSCode fork' tier money. These guys have built a fully built-out AI company with a range of models and applications.
525.
by torginus 6mo ago
afaik they serve some purpose in regulating androgenic-estrogenic hormone production. The amount of testosterone in women is not zero, likewise the amount of estrogen in men is not zero as well, and breast tissue does serve some purpose in
526.
by torginus 6mo ago
I just had a company Thinkpad break because its fan started rattiling. It was just out of warranty, like 3.5 years old. Within a span of a year, out of my dozen coworkers who have the exact same laptop, half of them went down with similar
527.
by torginus 6mo ago
Let me give you and example of what I referred to - I used to work on a big enterprise app that was a desktop app, it had a dockable very complex user interface and a quite sophisticated data management layer that you were supposed to integ
528.
by torginus 6mo ago
I wonder if it would work if top US companies implemented a system like the NFL draft, where companies competing for top engineers out of college get to pick from the best engineers inversely proportionally based on how they did before fina
529.
by torginus 6mo ago
Personally my experience has been that once I manage to describe a problem in good enough detail that a junior engineer would be able to solve it, it's good enough for an LLM as well. Which creates incentives I'm not wholly comfor
530.
by torginus 6mo ago
I have switched to OpenWRT during the LLM era. I wanted to set up some special network configs, and ChatGPT happily spit out the necessary configs. From what little I understood from OpenWRT everything looked fine, but nothing worked. I sti
531.
by torginus 6mo ago
I dunno if cycle length is the key here, the Soviets and the Chinese went with five-year plans, and done properly, it seems like thats a long enough amount of time to accomplish very important things. WW2 took slightly less than 6 years, wh
532.
by torginus 6mo ago
I think it's an important property of a system to be document able not just document ed . What I mean essentially, is the system was designed with sound principles, and said principles were written down and followed. I have seen this
533.
by torginus 6mo ago
Tbh, a ton of academic papers are quite poorly written. I'm not a PhD researcher, but I did have to implement quite a few of the, (computer graphics, signals & systems etc), and with most of them, I basically reconstruct the author
534.
by torginus 6mo ago
I'm kinda baffled by this whole belief system, that instead of seeing that other guys on the other side of the planet have managed to do what is generally though to be the pinnacle of Western engineering & investment with the fract
535.
by torginus 6mo ago
The content we see here is that you are Swedish. I am not sure what sort of moral, technical, financial authority are we supposed to be deriving from this. Dont get me wrong, Sweden is a cool country, but still my point stands.
536.
by torginus 6mo ago
I see, thanks. I guess most current models are not yet trained for this loop. Could you please try with Opus 4.7? I think there's a chance of it doing well, considering the design/vision focus.
537.
by torginus 6mo ago
This is just a random thought, but have you tried doing an 'agentic' pelican? As in have the model consider its generated SVG, and gradually refine it, using its knowledge of the relative positions and proportions of the shapes ge
538.
by torginus 6mo ago
https://en.wikipedia.org/wiki/List_of_most_popular_social_pl... There are all kinds of social media, its segmented by userbase, and culture/geography. Telegram has 1B users (which is surprising to me, I thought it
539.
by torginus 6mo ago
I know some pro artists (ppl doing work for big name companies, games, US film studios), either on a contract or employed basis. They've always told me the same thing - the job is to hit the minimum acceptable level of quality (which t
540.
by torginus 6mo ago
From what I can tell, its more about cashflow - basically companies need to spend most of their revenue or be taxed on it - and you can buy only so many servers. Now capital can flow towards AI - I'm sure the reason why engineers at Bo

More ›