Mars008

13 ms·

151.
by Mars008 1y ago
It's a tool which is like others. It has limits and needs to get used to. From my experience first attempts are overoptimistic. I tried big chunks at once. Then I learned smaller tasks work better. But the most important thing is it ca
152.
by Mars008 1y ago
To begin with you can get a cheap robotic kit from Amazon, there are many of them, and put it together. That's probably the easiest and fastest way. From here you can read more about servos and controllers, modify its mechanics and sof
153.
by Mars008 1y ago
> “The results surprised us,” research lab METR reported. “Developers thought they were 20pc faster with AI tools, but they were actually 19pc slower when they had access to AI than when they didn’t.” They forgot about training and pract
154.
by Mars008 1y ago
As I understand it they are saying users waste significantly less time on search if there are AI overviews. Is it really bad? Not from users perspective.
155.
by Mars008 1y ago
That's why I don't use mobile for browsing. Only Firefox with JS disabled unless it's needed and I need that page.
156.
by Mars008 1y ago
It still makes sense to use external data storage for smaller local models. They just can't hold that much.
157.
by Mars008 1y ago
One problem is in datasets which use RAG. Training foundational model requires a lot of samples, and there aren't many. The only option is to use other models to generate, so called distillation. BTW, RAG is similar to web search. Mode
158.
by Mars008 1y ago
> Seems pretty crappy to blame the websites for this instead of the AI companies, who are the really bad actors here Not sure who is bad. AI companies trying to make public knowledge easily available or web sites trying to lock and monet
159.
by Mars008 1y ago
> For example, Musk's obsession with colonising on Mars is troubling when you hear what scientists have to say about it. Musk tends to talk about it as if Earth is already lost. He is pushing the idea using all techniques. This one
160.
by Mars008 1y ago
So far latest 2 attacks are only in Ukrainian media and reprints
161.
by Mars008 1y ago
Can you buy Jules Verns book, add comments and claim copyright on the whole book? Claim partial copyright without specifying clearly what exactly?
162.
by Mars008 1y ago
Add to that it was done in the North where ocean's cooling is needed for Golf stream to keep circulating.
163.
by Mars008 1y ago
Whatever, that's doesn't justify shelling a city. BTW, it's considered Ukrainian territory and those civilians killed by Ukrainian military were Ukrainians. However it was supported by the government up to and including Zelin
164.
by Mars008 1y ago
> Ukrainians have not and do not deliberately target civilians There news on Russian side daily of Ukrainians doing exactly this, deliberately targeting civilians. Also usually their own non-combatants who are trying to surrender. This
165.
by Mars008 1y ago
> I keep a Windows 2000 virtual machine with no network access around just to occasionally play HOMM 3. according to wiki there should be an easier way: Platform(s) Windows, Macintosh, Linux (PowerPC/x86), iOS, Android Release March
166.
by Mars008 1y ago
For some of us this time has come, i.e. it's already useful and worth learning and spending on prime accounts.
167.
by Mars008 1y ago
> Russians are kiling civilians with drones each day. Same do Ukrainians, don't they? More over, long before they killed civilians and defenders in Iraq and Afghanistan on Americans invitation. Then left without even saying 'so
168.
by Mars008 1y ago
> each young person has to support more and more retirees, And here comes AI...
169.
by Mars008 1y ago
> Therefore Nvidia won't fail, because we're far far away from any AI production mode Not convincing, another cheaper and/or faster hardware producer can replace NVidia. Something like quantum or optical optimized for LLM
170.
by Mars008 1y ago
"The winner takes it all" economy. At least we have several AI providers (for) now. For short time it was only one.
171.
by Mars008 1y ago
> But its the LLM frenzy that made their market cap 20x in a few years, the other things didn't. Yes, and they have the second big wave coming. This time they have no close competitors. BTW, without LLMs Russian drones are using lat
172.
by Mars008 1y ago
AMD, Intel, IBM,.. were just as lucky, and what?
173.
by Mars008 1y ago
In other words we hate them and want them to fail. Bill Gates, Elon Musk, the same story repeats again. (c) NVidia isn't all about LLM, it has robotics, embedded, vision.. This is going to be huge as generic robotics hits.
174.
by Mars008 1y ago
I need both compute and memory. Video/image processing models take a lot in training. And the bigger the better at these sizes. So it will run trainings for weeks nonstop.
175.
by Mars008 1y ago
I couldn't find any video if it walking in real world. Looking at mechanical design I'm suspect it cannot walk reliably no matter how much you train it in simulator.
176.
by Mars008 1y ago
Irrational is what makes us different humans instead of one gray mass. History and traditions affect current values and perception of reality. And it's nice to have many of them instead of one 'common values' pack pushed by s
177.
by Mars008 1y ago
N100 more than enough for non-gaming desktop. It can play 8k 24fps 50MB/s x265 video smoothly. Which is impressive because 2080 RTX cannot.
178.
by Mars008 1y ago
Still have 2080 RTX on primary desktop, it's more than enough for GUI. Just got PRO 6000 96GB for models tuning/training/etc. The cheapest 'good enough' for my needs option.
179.
by Mars008 1y ago
Is AMD doing the same? From another post in this thread: > Nowadays, $650 might get you a mid-range RX 9070 XT if you miraculously find one near MSRP. If yes then it's industry wide phenomena.
180.
by Mars008 1y ago
I'm afraid google's search and AI bots are the same. Either get both or none. None means no visibility in web. Some will be fine with this, but most will agree on AI bots to get visibility. Soon AI providers will charge those who

More ›