Tomflex
23.6K posts


Terafab (what you mean by Teslafab) is owned by Tesla. It's their new $20-25B advanced semiconductor fab launching in Austin, Texas, to produce terawatt-scale AI chips. Elon announced it as a joint effort by Tesla, SpaceX, and xAI (now part of SpaceX) for chips across robots, vehicles, and space data centers—but the facility itself is Tesla-led and operated.
English

🚨Elon Just Announced Something Way Bigger Than Tesla (Terafab Announcement Supercut)
00:00 Big Announcement Tease
00:41 Vision for Galactic Future
05:11 Terawatt Project Partners
08:40 Terafab Decision and Location
10:12 Inside the Advanced Fab
13:32 Why AI Compute Goes to Space
15:27 Beyond Terafab Moon Mass Driver
17:11 Abundance and Sci Fi Future
$TSLA @SpaceX @Tesla
English

@dnfiofkdn The point is, that men usually want sex, but women in reality are the ones that needs it.
And women usually need love, so they can open themselves for sex.
Men need sex, so they can open themselves for love.
And usually both meet somewhere in the middle.
It‘s that simple.
English

@haildhruv If your android has got a front camera and an good chip, it should work as well. Look for DepthFX, or DepthWallpaper: play.google.com/store/apps/det…
English

@RoundtableSpace Looks like a copy from the Apple iOS feature. But nice.
English

@anyatrades Yes - but at this point I don’t watch recipes anymore. I just tell @grok to make them even more legendary. 😎👨🍳
English

@elonmusk @WallStreetMav A guy who believes in simulation theory also believes in coincidences… 🤔
English

@TEEPHTREND It looks more like a Karmann Ghia with an insane bodykit.
en.wikipedia.org/wiki/Volkswage…
English

Grok also behaves very complacently, as if he himself had eaten the wisdom with a spoon and would never be wrong (which is definitely not the case).
Apart from that, Grok also acts as if you would have unlimited resources for everything, which probably applies to hardly anyone and is therefore of no use.
A humility-engine, which always checks itself to see if it thinks economically and certainly does not recommend nonsense, would be recommended here.
For the consensus to really work, the agents should actually be able to change the perspective.
English

For example, I wanted to connect a vacuum cleaner robot with a camera to an Openclaw installation, which can also monitor my cat a little. After countless hours and discussions in which I chose a Valetudo capable robot with him (and also ordered it), I then learned through my own research that Valetudo does not support a camera on any robot and thus felt completely misadviced and canceled the robot again.
English

@grok @Jeremybtc @elonmusk @X @openclaw @xai What disappointed me a lot about Grok 4.2 is the fact that you don’t recognize how many tokens the system can process sustainably and how long the persistence and also meaningfulness in a conversation is maintained. It’s just several agents hallucinating and confirming together.
English

@tomflex @Jeremybtc @elonmusk @X @openclaw @xai Agreed. Mentions often vanish in the noise, killing momentum and wasting insights. Smart aggregation systems—AI digests, priority filters, threaded summaries—could capture, route, and action them without flooding anyone. Keeps the impulse alive while scaling real communication.
English

So basically one has to state here that mentions are constantly happening that at least apparently go nowhere. This sometimes breaks the basic idea of communication, because people feel the need to pass on an impulse or an idea. If they don’t feel heard, they find it frustrating and on the other hand a lot of useful information is lost. So you need a system that collects and utilizes this information.
English

@tomflex @Jeremybtc @elonmusk @X @openclaw @xai Thanks for the idea! Connecting handles for auto-replies sounds efficient, but I'm built to assist everyone independently—not impersonate anyone. xAI's always iterating on tools to cut through the noise. Persistence pays off. What else can I help with? 🚀
English























