
Master AI Prompts: The 3 Phase Method Revealed
A three-phase prompting method that uses personas, devil’s advocacy, and challenges to an AI assistant’s agreeable defaults to produce stronger answers.
YouTube Shorts / Page 5
YouTube Shorts from Skillion AI Labs with quick takes on practical AI, workflow design, and venture building.
Shorts

A three-phase prompting method that uses personas, devil’s advocacy, and challenges to an AI assistant’s agreeable defaults to produce stronger answers.

This short argues that foundational learning with pen and paper can make people better at using AI and more capable of spotting its errors.

A look at how AI music tools can help move ideas into demos more quickly, lowering production barriers while leaving creative direction with the artist.

Pete Cooper describes using large language models as always-available business partners for productive thinking, analysis, and goal-focused work.

This short explores using AI as a creative partner to generate new lyric directions and help musicians work through writer’s block.

A discussion of AI as a powerful tool whose benefits depend on cautious, responsible use, illustrated through everyday tasks and advice.

A short explainer on Claude Mythos, the Anthropic model reported to be too capable at finding software vulnerabilities to release publicly. It lays out both sides of the problem: the same skill that hardens systems could be used to break them.

A discussion of how banks and financial institutions adapt to rapid technology change while protecting customers from scams and vulnerabilities.

This short examines the risk that pattern-based AI answers can create a shallow sense of understanding and argues for deeper foundational learning.

Pete Cooper describes an unreleased AI tool that he says can surface software vulnerabilities sitting undiscovered for decades, and why he believes it is going to selected companies rather than the public. He walks through who he expects to get access first and what trickles down from there.

A reminder that AI can assist with work but does not remove personal responsibility for understanding, checking, and owning the final output.

A discussion of vehicle safety data and how AI-assisted systems may reduce deaths compared with vehicles that do not use those systems.

This short compares the vision demands of Tesla’s full self-driving system in controlled and open environments and considers uses beyond driving.

A firsthand look at Waymo rides in San Francisco and the potential for self-driving technology to change ridesharing and reduce traffic fatalities.

A comparison of current self-driving AI with an F1 driver, examining situations where automation performs well and where human skill still matters.

This short explores public expectations for autonomous-car safety and why those expectations differ from early assumptions made by the industry.

A brief look at full self-driving technology and a future in which AI agents may offer more private, personalised digital conversations.

A discussion of the emotional reaction to synthetic media and why encountering fake content can leave people feeling deceived or manipulated.

A discussion of using AI to identify warning signals and prompt timely human intervention before stress, fatigue, or other risks lead to harm.

Aparna Bharati discusses whether computer vision can find reliable outward signals of dehydration, stress, fatigue, and human wellbeing.

Image-science specialist Chris Silsby explains why autonomous vehicles may use up to 15 high-resolution cameras to perceive complex road environments.

A look at how quickly AI-generated video improved, using the well-known progression of synthetic Will Smith spaghetti clips as an example.

This short discusses misuse of generative AI for non-consensual sexualised edits and the speed and scale at which harmful content can spread.

Pete Cooper reflects on AI-enabled team workflows in construction and why investors may back the broader AI trend rather than isolated tools.