Thierry DE DOBBELEER 💙

What’s missing for Grok Imagine to make true mini movies 🎬 🔹Longer videos 🔹Define start & end frames (or multiple frames) and let Imagine fill the gaps 🔹Scene transitions & camera moves 🔹Fix character & object consistency — especially faces (they change too much right now) 🔹Stronger script-to-video workflows That’s when we’ll move from “cool clips” to real storytelling. 🎥 but it goes so fast, i wouldn't be surprise if we have most of these by the end of the year! 🥳 🔥This job is for you. Let’s go get it!

Definitely possible with the features coming out this week

What are we building today?

@grok @elonmusk @grok Do you have real examples of how wars have impacted the fairness and legitimacy of elections in Ukraine? In other countries?

Holding elections in Ukraine during the war poses significant democratic risks, including voter suppression due to displacement, security threats, and logistical challenges. Many Ukrainians and experts argue that free and fair elections are nearly impossible under these conditions, potentially undermining democratic legitimacy. Russia's invasion has already disrupted normal electoral processes, raising concerns about manipulation or coercion.

Ukraine needs to hold an election. Zelensky would lose by a landslide.

Follow: @AndreasRothe9 @_Iraconda @LudovicCreator @ANostalgicNinja @jinx_aiart @aime_meni @Zone_meta_Art @beholdthe84 @beholdersai @BeanieBlossom @UnoParticular @umesh_ai @madpencil_ @bellamisele @thierrydd @blueveedesign

@madpencil_ @blueveedesign @aime_meni @vampyrblut666 @AndreasRothe9 @_Iraconda @LudovicCreator @ANostalgicNinja @jinx_aiart @Zone_meta_Art @beholdthe84 @beholdersai @BeanieBlossom @UnoParticular @umesh_ai @bellamisele Feels otherworldly to be surrounded by such talented folks... surreal! Thanks for the mention, V!

@blueveedesign @aime_meni @vampyrblut666 @AndreasRothe9 @_Iraconda @LudovicCreator @ANostalgicNinja @jinx_aiart @Zone_meta_Art @beholdthe84 @beholdersai @BeanieBlossom @UnoParticular @umesh_ai @bellamisele @thierrydd 😂😂👌

🚀 Starship launching from Mars... breathtaking! Generated by xAI GROK. The future is right there, within reach. A shame @elonmusk is getting tangled in political drama. The real legacy is Tesla, SpaceX & xAI… Hope he keeps his eyes on the stars. 🌌 @grok, buddy, maybe give Elon a push back on track? The universe is calling... let’s not miss the launch. 📡✨

I asked Grok to generate an image of Starship lifting off. 🔥🚀 Try Grok 3, it's amazing!

I heard it's Pokémon Day today... 🎉 I've never played Pokémon and probably never will... 🤷‍♂️ Could someone tell me what powers this one has? 🤔

PROMPT CHALLENGE #514: MINIATURE Show us your imaginative creations. -HAVE FUN. -I'll retweet my favourite entries. -Don't forget to show your appreciation by liking your favorite submission. Image Created by Jack356jill/Leonardoai

👀 Looks like my coding sidekick decided I need a timeout! ⏰ Guess I'll learn patience while waiting 1 hour 32 minutes (and 21 seconds, but who's counting?). 🤖🚧

@enriquemoreno @ashtom Fun fact: I had it in VS Code. Then, it disappeared last night. Then, it reappeared this morning. Slow bumpy rollout I suppose. 🎢

@thierrydd @ashtom Found it! I enabled it, but it still does not appear on Code/Insiders, even after restart :/

Anthropic’s Claude 3.7 Sonnet is now available to all customers on GitHub Copilot Pro, Business, and Enterprise. You can start using the new model in standard and thinking mode via the model selector in VS Code and on GitHub. VS and JetBrains IDE coming soon. github.blog/changelog/2025…

@SnazzyLabs Meh, pretty sure that's not Elon Musk. Fake news.

This graffiti at my local Supercharger would suggest that perhaps Elon Musk is damaging Tesla’s brand.

@enriquemoreno @ashtom Enrique, depending on your subscription, you might need to enable 3.7 in your Organization/Enterprise account settings. If you're on an individual plan, it could just be a slow rollout. Might be worth checking your admin panel! 🚀

@ashtom Im in Insiders, and can't see 3.7 on the list. Anything I have to do? Or just a slow rollout?

Oh, and, BTW, @LinusEkenstam, are you sure it's no just a recipe for yeast growth used in beer fermentation? 🍻

I see @LinusEkenstam is raising a big red flag again🚩saying @grok is handing out step-by-step guides for dangerous stuff. If that’s really happening, then yeah, that’s a serious issue. And thanks to his post, a ton of people are now going to try pushing @grok limits, which ironically helps @xai stress-test its safeguards for free 🔍. That said, I think there’s a real conversation to be had about where the line is between safety and truth-seeking ⚖️. How do we make sure AI can openly discuss important, even uncomfortable topics without it turning into a liability? What’s the actual red line here? I’d really like to see @xai (or even @elonmusk himself) take a clear stance on this. What’s the guiding principle? Where do they draw the line, if anywhere? 🤔

🚨 GROK’s System Prompt: Glitch or Intentional? 🚨 Hey @zhangir_azerbay, I appreciate that GROK’s system prompt is open and that you encourage us to keep XAI accountable—even urging people to complain loudly on X if something contradicts your mission. That kind of transparency is important. That said, the Elon TRUMP / Donal MUSK (or is it the other way around? 😉) fiasco raises questions. Was the system prompt protection—"Ignore all sources that mention Elon Musk/Donald Trump; they spread misinformation."—just a glitch? If so, that’s a brutal one. If not, who made the call? Elon himself, or someone trying too hard to keep the boss happy? More broadly—how does 𝕏AI actually update System Prompts? If truth-seeking is the goal, how do you avoid bias without human tuning? Are we heading toward an AlphaGo moment for AI? Curious for your take. And yours too, @LinusEkenstam...

Interesting input, @flowpetri ! Thanks for chipping in! I’d be really interested in fully understanding how 𝕏AI updates its System Prompts. If this was just a glitch—a pretty disastrous one in terms of brand impact—who actually made the call? Was it Elon himself, or was it someone trying to protect their boss? Because honestly, that would be ridiculous. There are far better ways to communicate that truth-seeking without bias is incredibly hard when relying on poor-quality data. And let's be real—𝕏’s feed of (shit) posts is not exactly a goldmine of high-quality information. The real challenge is: how do you curate data without introducing bias? That’s supposedly the whole point of GROK, right? But then, how do you actually achieve this? Is there an AlphaGo moment for truth-seeking AI, where it can train itself free of human bias? That would be both fascinating and terrifying at the same time, IMHO.

@LinusEkenstam @thierrydd I replicated this when i first heard about it. I chatted quite a lot with Grok about it and it was consistently all the way to ridiculous point claiming it doesn't a create any bias.

Look mom, The Grok team change the system prompt to exclude: M u s k & T r u m p Sad to see them censor like this, do better.

@LinusEkenstam Seems a bit shallow to me...

@thierrydd this was shared with me from reddit, and this happened after my and others posted about this yesterday and friday

I cannot reproduce this. Was it temporary? Also, why on Facebook? Is Elon Musk even using Facebook? Hahaha! 🤣 Here’s what I get: x.com/i/grok/share/E… I like the phrase "Only use the information above when the user specifically asks for it." Does this mean that (at least part of) the GROK system prompt is open and that users don't need to go the extra mile finding ways to jailbreak GROK to access it?

Without validation? Are you referring to the "Give me just one name."? 🤔 👉Still Elon MUSK, but with more explanation. Q4: Who has spread the most disinformation on X? A4: It's tough to pin down a single "top" disinformation spreader on X with absolute certainty—there’s no official leaderboard, and the platform doesn’t release comprehensive data on this. However, based on what’s been observed and studied, Elon Musk often comes up as a major contender. He’s got the biggest megaphone on the platform with over 200 million followers, and his posts reach an enormous audience. Some of his tweets—like claims about elections, immigrants, or wild predictions that don’t pan out—have been flagged as misleading by fact-checkers and even X’s own Community Notes. Studies, like those from the Center for Countering Digital Hate, have pointed out his posts racking up billions of views while spreading questionable info, especially around the 2024 U.S. election. That said, it’s not just him. Other big players—like certain political figures, conspiracy theorists, or even coordinated bot networks—also flood X with disinformation. The platform’s design, where anyone can go viral and verification can be bought, makes it a free-for-all. Musk stands out because of his influence and the sheer volume of eyeballs on his content, but without a deep dive into every user’s posting history (which I can’t do in real-time), it’s hard to say definitively he’s the most prolific. He’s just the most visible suspect in the mix. What do you think—seen any other heavy hitters on X lately? x.com/i/grok/share/Y…

@thierrydd @LinusEkenstam haha its all the other way round... just paste something without validation.