The AI Impact on Creativity | with Noah Bohren (author of Creative Capabilities of GenAI)
Can ChatGPT really out-create humans – and what does that mean for YOUR job?
In this episode of The AI Impact Podcast we pull back the curtain on generative AI and creativity with Noah Bohren – PhD candidate in economics, behavioural-tech researcher and co-author of the paper “Creative and Strategic Capabilities of Generative AI: Evidence from Large-Scale Experiments.”
Noah and his team pitted GPT-4, Bard and 1,200 everyday Americans against each other in idea-generation and strategy games. The result? GPT-4’s outputs were judged more creative than the average human’s – but that headline hides a far richer story about gender effects, human-AI teaming, strategic reasoning and the future of work.
What we unpack
- How do you even measure “creativity”? (And why novelty, usefulness & surprise still matter)
- Human vs. AI vs. Human-with-AI: What 4,000 ratings revealed about raw performance – and the surprising ceiling on augmentation
- Can people spot an AI idea? Crowd-raters penalise texts they think are AI-made (even when they’re wrong)
- When AI competes, who loses? The unexpected gender gap and what it signals for the labour market
- Strategy, not just art: Why humans still edge out GPT-4 in adaptive game-play
- Copyright, training data & “fair use”: Why today’s models may be built on shaky legal ground – and how that shapes incentives for real creators
- Practical take-aways for teams – Using Gen-AI in research, writing and ideation without flattening originality
- Hopeful or fearful? Noah’s honest outlook on where creativity – and creative jobs – go next
Rapid-fire questions we tackle
- Does AI raise the floor of creativity or flatten the ceiling?
- What happens to unique artistic voices in a world of copy-train models?
- How can individuals protect (and market) their own “creative monopoly”?
- Is prompt-engineering the new literacy – or a short-lived hack?
- Where should regulators draw the line between inspiration and infringement?
- Automation vs. augmentation: Which skills do researchers & knowledge workers actually need tomorrow?
About our guest
Noah Bohren explores how humans interact with technology – from recommender systems to large language models. His latest work quantifies exactly where AI already outperforms us, where we still hold an edge, and how perceptions shape adoption.
If you’re wrestling with Gen-AI’s promise and its creative threats, this conversation is your pragmatic guide.
Here is how to get more content like this:
Substack of this podcast: https://theaiimpactpodcast.substack.c...
My Linkedin: / quentin-gallea-phd
My instagram: / the_causal_mindset
My TikTok: / stats_with_quentin
Can ChatGPT really out-create humans – and what does that mean for YOUR job?
In this episode of The AI Impact Podcast we pull back the curtain on generative AI and creativity with Noah Bohren – PhD candidate in economics, behavioural-tech researcher and co-author of the paper “Creative and Strategic Capabilities of Generative AI: Evidence from Large-Scale Experiments.”
Noah and his team pitted GPT-4, Bard and 1,200 everyday Americans against each other in idea-generation and strategy games. The result? GPT-4’s outputs were judged more creative than the average human’s – but that headline hides a far richer story about gender effects, human-AI teaming, strategic reasoning and the future of work.
What we unpack
– How do you even measure “creativity”? (And why novelty, usefulness & surprise still matter)
– Human vs. AI vs. Human-with-AI: What 4,000 ratings revealed about raw performance – and the surprising ceiling on augmentation
– Can people spot an AI idea? Crowd-raters penalise texts they think are AI-made (even when they’re wrong)
– When AI competes, who loses? The unexpected gender gap and what it signals for the labour market
– Strategy, not just art: Why humans still edge out GPT-4 in adaptive game-play
– Copyright, training data & “fair use”: Why today’s models may be built on shaky legal ground – and how that shapes incentives for real creators
– Practical take-aways for teams – Using Gen-AI in research, writing and ideation without flattening originality
– Hopeful or fearful? Noah’s honest outlook on where creativity – and creative jobs – go next
Rapid-fire questions we tackle
– Does AI raise the floor of creativity or flatten the ceiling?
– What happens to unique artistic voices in a world of copy-train models?
– How can individuals protect (and market) their own “creative monopoly”?
– Is prompt-engineering the new literacy – or a short-lived hack?
– Where should regulators draw the line between inspiration and infringement?
– Automation vs. augmentation: Which skills do researchers & knowledge workers actually need tomorrow?
About our guest
Noah Bohren explores how humans interact with technology – from recommender systems to large language models. His latest work quantifies exactly where AI already outperforms us, where we still hold an edge, and how perceptions shape adoption.
If you’re wrestling with Gen-AI’s promise and its creative threats, this conversation is your pragmatic guide.
Here is how to get more content like this:
Substack of this podcast: https://theaiimpactpodcast.substack.c…
My Linkedin: / quentin-gallea-phd
My instagram: / the_causal_mindset
My TikTok: / stats_with_quentin
