"World-class individuals do not like to be around fuzzy people."
Didn't understand this before, now I understand it.
When someone in the team is slacking off, it affects my motivation to produce the best quality work.
Distant yourself from fuzzy people.
Previously in 2nd now in 1st. Am a workaholic who truly enjoys what I do. Worked all the time even when I was making 4k
Then lucky enough to find a company who values my work and pays more. Working same hours, no diff to me
Doesn’t rly change how I spend but nice to be able to treat myself once in awhile
Day 25/
We are doubling all usage limits for all Pro 100/200/500 tiers and Business Premium, effective today.
We heard your kind feedback and won’t be reducing usage limits for Pro 200, so your Work and Codex usage will stay the same. We’re also going to continue offering extended limits for dots for now, and usage of your dot will still continue to be separate from your default usage.
And we'll of course also keep shipping better models and other efficiency improvements that increase your effective usage over time.
I did say we'd be releasing our models periodically...introducing Gemini 4 Carbon, our next frontier model.
Following Argon's release just two weeks ago, Carbon shows the pace at which we’re pushing the frontier, it ranks #1 on the recently updated v5.0 of @ArtificialAnlys Intelligence Index at 59. It also leads across key cybersecurity evaluations scoring 88.2% on SEC-bench Pro for finding vulnerabilities, 54.8% on ExploitGym for exploitation, and 71.0% on CWE-bench for patching.
We’re seeing significant improvements in long-horizon tasks even compared to Argon, and we’ve strengthened frontier safeguards.
As we did with Argon, we’re going to roll it out gradually + responsibly. Some trusted cyber defenders will get access starting this week. We'll be expanding it to more partners through our Fairwind Program, ahead of wider availability soon, so hang tight!
@atmoio Here u go sir: https://t.co/cgaIMKSZgF
Fully edited with AI end-to-end for client work.
It takes a shitload of refinement prompts but it is doable.
U can still see blobs and timing mismatch but I have tried to minimize it
My 3 biggest takeaways from making this video using AI prompting.
1. Be very specific with camera angle and movement.
Learn the names of different types of camera movements (pan, tilt, dolly, orbit, etc), it will save your AI model from guessing what type of shot u really want. And tell it what is the subject of focus, it will track the subject much better.
2. Include a speed ban/threshold.
Motion immediately looks like slop when it's too fast. Opus 5.5 is very good at generating motion, but it doesn't know how to hold back on speed. Most motion graphics by Claude is WAY TOO FAST to comprehend the message that is being delivered.
So, set a speed limit for element movement (i.e. UI buttons, text, 3d figures) and camera movements (you need a speed limit for each type of camera movement. tedious to configure but necessary). This limit can be done through test scripts that checks the code before the video is generated, or scripts that inspect the objects in the video after the video is generated.
3. Claude is still not great at 3D human/animal figures that MOVE.
70% of the time and tokens spent on this project is waiting for it to refine the surfer/skater/knight/horse.
Redid many times due to awkward or stiff human poses, or body parts disappearing/overlapping (you can actually see this happening if u micro-scroll the playhead and check frame-by-frame).
At some point it's diminishing return - spending an insensible amount of tokens for very little improvement. So I stopped at what I think is 80-90% acceptable.
_________________________________
Conclusion is, with very careful prompting, you can achieve very intricate animations and smooth motion that doesn't look like slop.
And I think prompting is going to be the new method of generating motion graphics, moving forward.
When smarter models get released, they will get better at understanding our intentions even with bad prompting, so in the future we'll probably be able to get far better results with way less baby-sitting (kinda like how it happened for programming too)
Another video where the motion graphics and animations were generated through prompting.
Tons of refinement prompts went into it to make it come out the specific way I want, and of course with proper SFX added manually to sell the quality.
Over the past year, the LP Army has grown into a global community of retail market makers.
From learning the craft to sharing strategies and onboarding new LPs, you've all helped Meteora grow.
To everyone who contributed, thank you.
More on Expert Contributors soon.