Is Seedance 2.0 a Killer? | ✉️ #88
Hey! 👋
Seedance 2.0 has dropped like a bomb online, and the internet is doing what it always does: going crazy with it. In just a few days, social media got flooded with clips, tests, remixes, and “how did they do that?” videos. It doesn’t feel like a normal product launch—it feels more like one of those moments where everyone suddenly realizes something big just changed.
What’s making people talk is that this isn’t just another AI toy making weird clips. The videos look way more polished, and people are getting much better control over the results—camera angles, movement, lighting, pacing, all that stuff. So instead of just typing prompts and hoping for the best, creators are starting to feel like they’re actually directing scenes. That’s a huge jump.
And the marketing side of this has been wild too. Seedance 2.0 is basically built for the social media era: creators test it, post the results, other people copy the prompts, and the whole thing spreads even faster. Tutorials, hacks, prompt packs, reaction videos—everything feeds the hype. You can almost watch the ecosystem growing in real time.
Meanwhile, the film world is definitely feeling the pressure. It’s not that AI is replacing Hollywood overnight, but it’s already getting good enough to shake up parts of the process: pre-vis, concepts, ads, promo content, and lower-budget production. And that’s usually how these things start—first it helps with “small stuff,” then suddenly it starts replacing expensive workflows.
So yeah, here we go again. Another AI tool shows up, everyone laughs for five minutes, and then the whole industry starts panicking. Seedance 2.0 feels like one of those turning points where the rules start changing again. And once more, AI may have just pushed the timeline forward before anyone was really ready for it.
What We've Shared
- THE BEST OF DEVOPS ACCENTS, volume 1: Here's a little compilation of some of our best clips from past episodes!
What We've Discovered
Running Claude Code dangerously (safely): With AI agents running on our machines, we are going to see another spike of VMs and containers adoption - a bit cheaper that getting a MacMini just for this purpose!
The future of software engineering is SRE: Compelling (for us as a consultancy in this space) argument, that infrastructure engineering and SRE are going to be become more important with the ongoing explosion of AI-generated applications.
Using the shared plan cache for Amazon Aurora PostgreSQL: Using less memory also means paying less for RDS. The only question that remains unanswered is - what are the potential pitfalls?
How AWS S3 is built: Lots of fascinating details about the 8th wonder of the world.
You Need More AWS Accounts Than You Think: An overall good baseline, though an explanation of why split Artifacts from CI/CD account is needed.
The 89th mkdev dispatch will arrive on Friday, March 13th. See you next time!