I personally agree, but (some) people stop carrying their wallets, when they can pay with their phones.

Became is a bit generous

I’m not seeing it as learning as behind the scenes the questions are changed, instead of the answer to the same question is becoming correct.

Also it becomes rather severely limited in the context length, or in this case in how much can be “learned”.

but these are still… prompt extensions (not sure if there is a technical word for it), right?

that’s a neat workaround for context windows, but at the core, imho any intelligence must be able to learn, and for a neural net to learn, it must change the network, i.e. weights or connections.

does it matter how light MacOS is, when nearly everything is in the web, which is very much not light?

don’t LLMs generally already fail at the learning stage of Intelligence?

once trained, they never learn again? It just sometimes seem like they are learning, as long as the learned thing is still within their “context window”, so basically it’s still within their prompt?

In another matter, how would we evaluate actual intelligence with LLMs? Especially remembering that all of the slop-companies would immediately try to cheat the test.

Well, he could have backed out. (Which would have been safer)

With this ruling FIA says that that’s not required.

Depends, index alone (no trackers, no controllers) is Quite a bit cheaper

I personally expect that they will want to cover hardware cost, while being willing to subsidize development cost with the steam store.
So I think the low storage models will make a slight loss for valve in the beginning, while the higher storage models will get that back to even overall for valve.

I’m extremely curious about the price.

For a long time, there were $1200 rumors.
Now we have the “less than index”, which I believe spawned the “under $1000” rumors.
but since index has a huge range, depending on the accessories you buy with it “less than index”, can mean anything from <$600 to <$1100.

But in reality this has to compete with Quest 3, and preferably also with Quest 4 when that releases.
So I think it really should be on the lower side. <$600 would be good, <$500 would be great, <$700 would be okay

The most disappointing part is the rumoured pricing of “aiming to be under $1000”.

$1 is “under $1000”.

at $999, even for the 1TB model, this is a really tough sell.
I’m not sure I’d get one, even though I love what they’re doing, and want to support it.

I really hope, that the “under $1000” is a misunderstanding from the “cheaper than index”, which currently sells for 539€ (~$625, incl tax) without controller and base stations.
That would be a great price.
I can just hope it’s nearer to that than the $1000.

I think all of these are nice, if priced correctly

Steam Frame needs to compete with quest, so prices over $800 are a really tough sell.
Steam Machine needs to compete with consoles, PS5 (non-pro) and Series S, so prices over ~$700 will become really tough.

Prices start becoming really good, if they manage to come it at ~$600 for Steam Frame and ~$500 for Steam Machine.

But with current hardware prices, Valve being valve and no-one can know if they want to make money on the hardware, or if they are willing to sell at cost, or if they are willing to subsidize, who knows where we will land.

Prusa is way more open, but significantly more expensive, especially when buying assembled.
If youbwant multicolor/multimaterial their current (fairly soon to be replaced) solution is not considered as user-friendly as the current bambu-solution.
Yes, when the build volume is 10x10x10 you can print things within that volume, but of course it still has to be a printable shape.

A T shape for example would be difficult to print, printer print layer by layer and as the “Arms” on the top would have nothing to be “stuck on”, so you’d need what is called “supports”, a printed shape just there to support the actual object that you want to print. Usually were support meets object the surface quality of the print suffers to some degree.
In the case of a T shape, just print it upside down then ;)

I just want to note that they recently did take functionality away. Home Assistant Integration and Panda touch and orca slicer for example, all of those have to either use lan mode, not update the firmware or jump through more hoops than before

Gameyfin exists as well

Not even that though, just that his opinions are not always wrong, and his investments sometimes right

Well, reusable rockets are worth it, LEO satellite Internet being worthwhile, electric cars being a thing.
Stuft like that

You do realize that you lose quality with wach encode, right?

It’s not AS bad when bitrates are high, but it’s still there.

But that’s exactly the point.
if the virtual map they’re building from cameras is complete, correct and stable (and presumably some other criteria that I didn’t think of from the top of my head), then the cameras would be sufficient.
The underlying neural decision network can still fuck things up from a correct virtual world map.

Now, how good is the virtual world map in real world conditions?

Well, do we know what the blockers are for Tesla?
I feel like when I watch videos of FSD on cars, the representation of the world on the screen is rather good.
Now given this datapoint of me watching maybe 30minutes of video in total, is the issue in:
a) creating the distance to obstacles in the surroundings from cameras or in:
b) reading street signs, road markings, stop lights etc, or in:
c) doing the right thing, given a correct set of data about the surroundings?

Lidar / Radar / Sonar would only help for a).
Or is it combination of all of them, and the (relatively) cheap sensor would at least eliminate a), so one could focus on b and c?