One of the most exciting and challenging parts of working with media companies on their AI strategy is keeping pace with how rapidly the tools evolve. In this space, “what doesn’t work” can become “industry standard” in just a few short months. Tools that once gave you mediocre—or even outright disastrous—results may have quietly gone through three or four major improvement cycles. And let’s be honest: after getting burned once, it’s totally fair for time-strapped teams to be cautious about giving it another go.
But the speed of improvement is real—and it’s happening across all types of media.
Take for instance the evolution in image generation. Here’s a side-by-side of a 2023 prompt showing a moose inside a CVS and the same prompt in 2025:
In under two years, the evolution of the image generation has gone from a bad Bullwinkle knock-off to one that is getting close to looking like it was taken by a late model iPhone.
For another example, this was posted to Reddit this week. Same prompt, two years apart. Click on the image to see a video of the otter opening the laptop!
And that’s just visuals. Text, audio, and video AI tools are also leveling up dramatically. What might have felt clunky, robotic, or just “not ready for air” in 2024 could now be polished, broadcast-worthy, and a huge time-saver.
So if you or your team ran into a wall last year, maybe a voice sounded a little too uncanny valley, or a generated clip didn’t pass the muster, give it another shot. The tech that let you down may have quietly become your late-Friday-deadline savior.
In short: Don’t base your 2025 decisions on 2023 tools. Re-test. Re-evaluate. And if you’re not sure where to start, I’m happy to help you navigate what’s working right now.
YouTube Adds AI Music for Creators
YouTube has introduced a new AI tool called “Music AI” that lets creators generate original background music simply by typing in a prompt. This tool, currently in testing with select creators, can produce royalty-free tracks in various moods and genres, tailored to the video’s theme. It’s designed to make content creation faster and easier while minimizing those pesky copyright violations. [details]
No Fakes Bill Gets Reintroduced in US Senate
The entertainment industry is rallying behind the reintroduced bipartisan “No Fakes Act,” aimed at curbing unauthorized AI-generated replicas of individuals’ faces, voices, and names. Supported by major platforms like YouTube and organizations such as SAG-AFTRA and the Recording Industry Association, the bill allows individuals to request the removal of such content while limiting platform liability if they act promptly. [details]
Recommended Read: The rise of deep research: How agencies are using AI for strategy, content and client pitches
Defendant Scolded For Using AI ‘Lawyer’ as a Court Representative in His Appeals Case
A 74-year-old AI entrepreneur tried to sneak an AI-generated avatar named Jim into a New York courtroom to argue his case—and the judge was not having it. As the case began, the man asked to play a video. When the video started playing, the judge asked, “Who is this man?” and the defendant casually dropped, “Oh, that’s just my digital avatar.” Cue the judge’s epic clapback: “You are not going to use this courtroom as a launch for your business.” [details]
Originally published by Jacobs Media
