Evaluating the impact of AI Agents, one job at a time.

In our previous post, we evaluated the role of Agentic AI on the role of Multimedia Artists & Animators. With an AI2 score of 77.19, it is one of the jobs that is of moderate-to-high risk of getting agentified. We also highlighted that the roles are not being replaced immediately, with a hybrid model where AI handles routine animation tasks, allowing human artists to focus on higher-level creative challenges: developing new storytelling approaches, exploring experimental visual techniques, and pushing the boundaries of what animated content can achieve.

Let us do a reality check on the capabilities we called out in our previous post. I have used both free and paid versions of some of the options. Please note that I am not an expert at this, check out links such as this for more sophisticated demos. This post is also not meant to provide an exhaustive list of SOTA market options.

  • Generate complete character animations from simple text descriptions

Simple Animation from text — created using SayMotion (DeepMotion)

Simple Animation from text, variation — created using SayMotion (DeepMotion)

While the second variant above is a bit closer to the prompt, both are far from satisfactory. May be need to get better at prompting.

Simple Video from text — created using Google Gemini Pro 2.5/ Veo3

Simple Video from text, closeup version — created using Google Gemini Pro 2.5/ Veo3

Satisfactory experience, however, 8 seconds limit with the Pro plan on the duration does not allow generating any meaningful videos. Probably the Ultra plan allows longer videos?

  • Create photorealistic motion capture from basic webcam footage

Motion Capture from Video — Using Animate 3D (DeepMotion)

  • Produce studio-quality lip-sync animations in real-time

Avatar and Lip-synced Video created using HeyGen

Avatar and Lip-synced Video (verion 2) created using HeyGen

Ignoring the over-bite, both variations demostrated excellent lip-syncing. including for the occasional stumble and puases.

  • Apply sophisticated lighting and rendering effects without manual intervention

Lighting Effects — Using Adobe Express

  • Transform 2D drawings into fully-rigged 3D characters automatically

I am super impressed with Google Gemini’s ability to convert a sketch into a 3D video using Veo3 model. See the output video from the sketch below.

A skethc from the popular Calvin and Hobbes series

Video generated from the above sketch using Google Gemini Pro 2.5/ Veo 3

Conclusion

The above examples provide a sneak preview into what is possible — while these are just simple examples, experts can make create more appealing and meaningful work. Interestingly enough, CBFC (the authority that certifies all movies getting released in India) has also approved a first AI-generated film for theatrical release in India title ‘Love You’. There are also other attempts (Naisha, Hanuman) to make longer AI-generated feature films.

Akin to the early days of cloud computing, the foundational blocks are in place, but additional work is needed to orchestrate these into meaningful workflows.

With what is possible already, and the pace of innovation in this space (and GenAI/ Agentic AI in general), our stand that the future of this role is one of symbiotic human-AI collaboration remains the same.

This analysis is part of our #100DaysOfAgenticAI series, examining how AI Agents are reshaping professional landscapes across industries. Follow along as we explore the future of work in the age of autonomous systems.