@1001TLtv Muito fixe!!!

@Kwebbelkop This one reminds me of a dream.

@1mf_dave @_akhaliq it's funny how all these cutting edge visuals have been done before. Halftonejs is at least 10 years old.

Playing with knob design and audio in front end design and Snap!

@lukas_m_ziegler It failed on KickStarter back in 2024. I like the concept but unless they got other funding it's dead in the water.

Fill a Super Soaker with electro-conductive black paint. The paint will help conduct electricity, potentially short-circuiting the robot's systems when you soak it. Plus, the paint can also block sensors, even if it doesn't get inside, giving you an upper hand. Easy, inexpensive, and effective! #RobotTakedown #DIYDefense"

@RaphaelDabadie @sama Unless gpt-5 is more than a model it won't ever be smart in the way humans are. Models can't feel or experience only infer. Confusing un-embodied LLMs problem solving abilities for human intelligence I like not understanding why cars can't cut grass and fertilize gardens.

Testing Operator's ability to work with block based programming. :,( As of today it's especially bad at doing anything that a grade 4 student might do.

@naren_66 @vitrupo @EMostaque Agreed, too much of most work is routine. That said, we don't need opaque tools but systems that empower our minds

@Newaiworld_ @kimmonismus Everyone has access to these tools. Best we aim at being the best we all can be instead of playing into the hype. We can't worry about the replacement of 500 midlevel programers teams. We need to worry about how to think and create in ever more powerful and flexible ways. ❤️‍🔥🧠

@rezmeram @kimmonismus Both are not wrong.

@Donkeyhead6 @Rainmaker1973 likely 1/3 then you got the bots and people who don't better.

@Donkeyhead6 @Rainmaker1973 Yup, the whole description is off. They are trolling for engagement.

@PirateKey @OxUniMaths The waterfall

@Izzyplante @rohanpaul_ai Don't get be wrong they are likely extracting these videos from their model but it giving back almost the exact same video is a sign of over fitting.

@levelsio @eloytoro @NickSarafa Might want to just get an upscaler while you wait for the backorder to clear in that case.

What we're looking at isn't generative. If anything it's doing a search on the training data essentially compressed into the model. This isn't to say that the same model couldn't also be used to actually do generation, but if you look at the video of the kid for example there's the normal tells.

@altryne Oh, some of us understand that there's a problem with the content that they are claiming is AI Creations. I need a better mic if I'm going to cover more of this. In the meantime, don't believe all the hype.

@altryne I don't believe all of their content is real.