Human-in-the-Loop
Where AI workflows meet critical thinking in communications
Bitcoin hit just over 110 today.
To The Moon.
We the team are working on thinking through comms workflows. I’m taking online courses to dip a toe in some coding basics. Please vote here if you would take a free course on how AI is transforming communications teams work and how to ethically and productively optimize workflows and outcomes.
Here are some of the questions we get asked and some of our thoughts. I’ll keep sharing if you find it useful. Tell me.
How does AI enhance media strategy and planning workflow, and what are its core benefits?
The Rundown Team: AI enhances media strategy and planning by automating labor-intensive tasks, and so increasing efficiency and freeing up human experts for more critical and creative work. The core benefits include quicker turnaround times for comprehensive research, more data-driven insights, the ability to process vast amounts of information efficiently, and the capacity to generate diverse content options. This ultimately allows for a more analytical and efficient approach to communications.
What is the "human-in-the-loop" approach, and why is it crucial in AI-driven media workflows?
The Rundown Team: The "human-in-the-loop" approach emphasizes that while AI performs extensive research and preliminary tasks, human oversight and critical thinking remain indispensable at every critical stage. Humans define the initial parameters for AI prompts, review AI outputs for accuracy and relevance, apply editorial judgment, and provide nuanced direction that AI cannot grasp (e.g., sentiment, cultural sensitivity, specific client history). This iterative process allows comms specialists like us to make sure that the AI's output is high-quality, on-brand, and aligned with the client's specific needs and ethical guidelines, preventing generic or inaccurate results.
Critical Thinking Please !
News
On software development: Reuters reports that AI coding tools like Microsoft's GitHub Copilot brought in over $500 million in revenue last year. Google's CEO noted that 30% of their code is now AI-generated, while Amazon saved the equivalent of 4,500 developer years using AI tools. These developments demonstrate how AI is becoming integral to the software development process, accelerating innovation while raising questions about code quality and developer skills.
On job market. The New York Federal Reserve reports that unemployment for recent college graduates in the US jumped to 5.8% in the first quarter of 2025, up from 4.5% the previous year. Companies are using AI to replace entry-level jobs, particularly in technology.
On advertising: Meta is wants to automate end-to-end advertising creation and targeting by late 2026. Under this vision, brands could provide product images and budgets and off go AI systems to generate text, video, and ad placements optimized for user location and context. This will create a need for oversight and brand consistency.
Thanks Perplexity AI for the above three stories.
Tool: Opus Clip
I wanted to share a tool called Opus Clip which The Rundown Team is using that’s changed our workflow and efficiency. It transforms long-form videos into short, engaging clips optimized for social media platforms (ideal for YouTube Shorts, Instagram Reels, or TikTok) using AI. One piece of content should be repurposed on many different platforms. What we like about this tool us that
We save time, we save cost.
The AI identifies the most engaging moments of the video
Short clips are automatically created.
Video reframing for various aspect ratios (9:16, 1:1, 16:9)
Captions are mostly accurate (not always)
Get a score that predicts potential performance.
The main drawbacks are that watermarked exports on the free version and some inconsistency in automatic clip selection. Sometimes the AI picks moments that feel disconnected from the main narrative.
For communications teams producing regular video content it’s a gift. Video production budgets are often limited but social media reach is critical. Even with just the free plan we were able to create hundreds of clips off 3 Embedded podcast episodes. The multi-language caption support also helps reach broader audiences.
Thinking Vs Reasoning.
New research shows limitations in reasoning models that could reshape how we build and deploy AI systems in emerging markets. But how much does this new research from Apple stand up? You can read about it here:
While these models demonstrate improved performance on reasoning benchmarks, their fundamental capabilities, scal- ing properties, and limitations remain insufficiently understood. Current evaluations primarily fo- cus on established mathematical and coding benchmarks, emphasizing final answer accuracy. How- ever, this evaluation paradigm often suffers from data contamination and does not provide insights into the reasoning traces’ structure and quality.
Some of Apple’s tests included four types of puzzles.
Tower of Hanoi: Moving disks between pegs with specific rules
Checker Jumping: Swapping colored pieces on a board
River Crossing: Getting people across a river with constraints
Blocks World: Rearranging blocks into target configurations
Medium author Ali Sandar says that “the discourse surrounding artificial intelligence frequently features claims of machines capable of “thinking” and “reasoning.” This narrative has been significantly challenged by Apple’s recently published research paper.” Here’s his analysis.
Meanwhile I’m waiting for Apple to either fix Siri for once and for all, and make it useful or come up with something of its own in this space that’s truly exciting. By that I do not mean another iphone with more camera stuff.
Zain’s Homework
Adding these Veo3 prompting tips to my homework. I was interested in the gaming opportunities here.
Shot composition: Specify the framing and number of subjects in the shot (e.g., “single shot,” “two shot,” “over-the-shoulder shot”).
Focus and lens effects: Use terms like “shallow focus,” “deep focus,” “soft focus,” “macro lens,” and “wide-angle lens” to achieve specific visual effects.
Overall style and subject: Guide creative direction by specifying styles like “sci-fi,” “romantic comedy,” “action movie,” or “animation.”
Camera positioning and movement: Control the camera’s location and movement using terms like “eye level,” “high angle,” “worms eye,” “dolly shot,” “zoom shot,” “pan shot.” and “tracking shot.
Film short…coming soon.