Grok’s new model is way more tickle-aware.

[Toesheldback]

No need to say “poke” or “scratch” when you animate a photo. The word tickle suffices. Also the lee and ler say the word constantly.

 

[stockingfeetrbest]

I noticed this too. I created one today and posted it in this section. Jesters tickling another jester. The hand movements and finger wiggling was perfect. They also seem to be taking good direction for verbal cues. I always put what I want them to say in "". There is still a lot of error with the wrong person speaking or the lee saying the lers line. Getting better though. There is also an option for 10 sec videos now just in case you weren't aware. It's in the same drop down box where you choose image or video.

 

[stockingfeetrbest]

No need to say “poke” or “scratch” when you animate a photo. The word tickle suffices. Also the lee and ler say the word constantly.

Are you using image instead of video and then picking one of the many pictures generated and turning it into a video?

If you are, try using video instead of image. You can keep hitting the redo button and get as many versions of your prompt as you want.

 

[Nicklaos]

I also used to specify that the ticklers "move their fingers",another probles was that the AI has problems telling left from right and also couldn't properly position tickling hands under the soles of the feet.

 

[Toesheldback]

I also used to specify that the ticklers "move their fingers",another probles was that the AI has problems telling left from right and also couldn't properly position tickling hands under the soles of the feet.

Exactly. It’s painful still sometimes.

 

[Nicklaos]

Exactly. It’s painful still sometimes.

Did some tests yesterday before hitting the daily rate limit,I saw the improvement in the tickling animations but still some problems that cause body parts morphing in a weird way,I'm using https://grok.com/imagine because right now is the best free option for both images and videos while https://labs.google/fx/tools/whisk can be useful when you want to merge multiple images and compose a scene with specific characters but it's very censored.

 

[stockingfeetrbest]

They are constantly changing how Grok works and usually not for the better. As of today it insists on adding theatrical music to every clip which totally ruins them. For the past few months I have had to specifically write "there is no music" at the end of every prompt. Now it just does what it wants. Very frustrating. I have also noticed that they are covering up the crotch area every time now while in the past I was able to get some that were completely bare and others that you could easily see through sheer nylons.

 

[Credo 420]

I noticed a while ago with one of the updates literally every person's voice sounded similar. With laughter etc...

Also all the different users like us that are putting in the prompts I'm sure that it learns with every prompt.

 

[Nicklaos]

They are constantly changing how Grok works and usually not for the better. As of today it insists on adding theatrical music to every clip which totally ruins them. For the past few months I have had to specifically write "there is no music" at the end of every prompt. Now it just does what it wants. Very frustrating. I have also noticed that they are covering up the crotch area every time now while in the past I was able to get some that were completely bare and others that you could easily see through sheer nylons.

Yeah my last tests gave bad results and so I quit for a couple of days,I'm starting to think that depends on how much work load the site is experiencing,also the latest news regarding AI in general aren't positive.