@BradleyKSherman@StevenBartlett The difference is that “AGI” is a broad, disputed forecast made by many people. Here, we’re talking about one man making essentially the same imminent-crash call four years in a row.
@nptacek It does not prove browser-based engines are ready to replace native C++ engines or ship modern AAA games at scale. Web tech is progressing quickly, but a tech demo and a full production ecosystem are completely different things.
Buying a laptop from a bricks-and-mortar store and turning it on is “too much effort” to set up?
I understand that most people will choose convenience, and cloud models will probably remain ahead of local models for some time. My point is that it should remain a choice, not become the prescribed future of computing.
We don’t remove the ability to drive because autonomous cars exist. We don’t stop people cooking because a machine can do it for us. Automation should expand our options, not remove our agency.
@ClaudeDevs It should work more like Codex, where it can automatically fall back to a lighter model and keep working once you hit your main usage limit, rather than just stopping completely and waiting for the limit to reset.
At what point do you think we’ll be able to confidently say humanoid robots are truly autonomous?
Not autonomous for a carefully demonstrated task, but able to put one in an ordinary home, give it a list of jobs, leave for the day and trust it to handle unfamiliar situations without a human supervising or remotely intervening.
Kind of, but not as part of the actual SuperGrok subscription. Cursor Ultra is still a separate product/subscription. There is currently a promo giving eligible SuperGrok Heavy users a free month of Cursor Ultra, which might be what you’re thinking of.
My point was more that I’d like a native Grok desktop GUI so I don’t need Cursor at all. I want the Grok Build capabilities wrapped in a proper first-party desktop project environment rather than having to use the terminal or another provider’s harness.
I think you’ve misunderstood what I meant. I know Grok Build already supports Git, worktrees, branching and structured iteration.
My point is the interface and workflow around it. Right now, if I want to use Grok Build properly on desktop, I’m essentially working through the terminal unless I pay for an additional product like Cursor.
What I’m asking for is a first-party Grok desktop experience that wraps all of this into a proper project workspace, similar to what Claude and Codex offer, without needing another subscription or third-party harness.
I’ve even seen people on my TL harness Grok through Codex just to get that kind of workflow. That’s the gap I’m talking about. The underlying capabilities are there; I want a native Grok interface that actually brings them together.
I don’t use Cursor because I already pay for SuperGrok, and I don’t think another paid product should have to fill the gaps if Grok is supposed to be a standalone platform.
I’d love to be able to remotely control Grok Builds from desktop, rather than having that functionality limited to the mobile app.
The terminal is solid, but what’s still missing is a proper native desktop experience comparable to what Claude and Codex offer. For larger projects and heavier iteration, that makes a huge difference.
I’m already seeing people use Codex as a harness for Grok through GitHub, which says a lot about the gap that currently exists. Proper project support with branching, combined with desktop control of Grok Builds, would go a long way towards closing it and make the whole platform significantly more capable.
This is beautiful. I’ve always loved this kind of art style and aesthetic. Is something like this transferable to a game engine like Unity or Unreal, or is it essentially just an animated image? I’m curious because it looks like it could be a really fun style to experiment with for a game project.
I learned pretty quickly that these “games” aren’t really games at all. As things stand today, you can’t just one-shot a prompt and suddenly have a game with the art direction, design, depth and consistency of a AAA title, or even an indie one for that matter.
I think this guy is just trolling for engagement, because anyone who has actually tried building something substantial with Claude, Codex or even Grok knows how superficial the initial output can be. Getting from that first generation to something genuinely cohesive can take hours or even days of prompting, iteration and manual work, depending on what you’re trying to build.
The art simply isn’t there either. If you want a properly designed world with consistent environments, maps, characters and assets, you need other tools, AI or otherwise, and you still need to actually understand how all of those pieces fit together.
So yeah, claiming you can basically prompt a AAA-quality game into existence is bullshit. It massively downplays the amount of work involved and feels more like engagement bait designed to make him look like he knows his shit than an honest representation of what these tools can currently do.
@afcstuff@Plettigoal What a ridiculous offer considering how inflated the market is. He's worth 70 million minimum, and we need squad depth for the league.
@IGN It seems like it’s going to be a longer-form presentation. Surely it won’t just be 10 to 15 minutes. I’m expecting more of a 30 to 60-minute deep-dive documentary-style showcase. If it’s just another trailer, I’ll be pretty disappointed.
It looks like Vini used us as leverage to get Real Madrid back to the negotiating table. At this point, we should go all in for Álvarez because we desperately need to strengthen our attack.
What I saw at the Aviva simply wasn’t good enough, friendly or not. We lacked any real cutting edge going forward.