A story about creativity
BCG’s findings about GenAI and creativity (in a business context, naturally)
We share many more GenAI use cases from the rabbit-hole
We’ve written at length about the need for GenAI to deliver outcomes in order to avoid being used as a lightning rod for polarised, philosophical debate on whether AI will be our doom, salvation, or otherwise. For those who - like us - have huge optimism for the future it is important to understand the other side of the argument.
Last time out we covered a number of fully baked GenAI use-cases which broadly fit under the summary: if it’s language work and some errors are tolerable, use GenAI. Since then we’ve been taking a look at the swift changes that have been made to the creative process (nb// a perfect example of language work where errors are tolerable, maybe even preferable). It reminded us of having read this, written by Nick Cave of Nick Cave and The Bad Seeds fame.
ChatGPT is fast-tracking the commodification of the human spirit by mechanising the imagination.
Vitriolic stuff. That said, it’s hard to disagree with the idea that GenAI content is replication rather than something novel. However we’re often reminded by The Pessimists’ Archive that new technologies (a) are not nearly as worrisome as they may first appear, and (b) serve to change the field on which the game is played on, rather than changing the game.
Remember when the PlayStation2 and MacG4 were national security concerns? Well, no one has successfully tampered with missile guidance systems using a PS2 as far as we know. We all survived the emotional harm of owning tamagotchi. As for music in the future: we’re pretty sure that music and songwriting - some of humanities oldest and most entrenched pastimes - will endure. The ways in which those artists compete for our attention will change as they always have done.
We can’t state the importance of outcomes delivered by GenAI enough: without them it’s all too easy to end up in the acceleration//doom debate spiral. As a result, we are always on the lookout for ways in which innovations can positively affect our businesses and the wider economy. In recent days we came across this BCG survey. It’s enticingly called How People Can Create—and Destroy—Value with Generative AI.
An important finding was that if you use GenAI, outcomes for ‘creative product innovation’ (ie// coming up with experiments and ideas derived from the immense training data that the AI is trained on) were much improved. This was particularly the case for bottom-half performers: a real life demonstration of the power of GenAI tools if used in a training context.
Regarding business problem solving (ie// multi-variable issues with a lot of humans, opinions and competing alignments in the mix) GenAI is counterproductive. No real surprise there. The conclusion here is: we wouldn’t be advocating trust in ChatGPT to produce your 2024 strategy deck….but we’d encourage you to use it to come up with ideas for it.
Perhaps just as interesting is that in the areas that GenAI performed well, tampering with the output negatively affected quality. That human tendency to tinker, to put your own spin on things is already counterproductive. There are areas where GenAIs are better than us, we should take that as a luxury.
There’s a cautionary tale about creativity here however.
Creativity, like any skill, must be practiced. As we mentioned above, GenAI has changed the field which the game is played on. It can be relied upon in some areas, totally unreliable/undesirable in others, and a co-pilot for everything in between. Understanding this new playing field and regularly exercising our own creativity as an input into these processes will result in the best outcomes.
As you can tell, we’ve been down the rabbit hole on GenAI use-cases and wanted to share some that we’ve found on our travels which are specifically related to this week’s topic of creativity. As you’ll see, they all fall under our Occam’s razor of: is it language work and some errors are tolerable.
Reading and Writing
Record yourself talking, turn the monologue into a skeleton blog post
First pass research assistance, feedback on drafts
A creative muse: “Give me 10 headlines”
Powerful summarisation
Drafting standard operating procedures for back-office and administration such as: payroll, finance, HR (where you probably already have all of the assets required)
Ideation
Give me 10 examples of an something, rank those, request 20 more of the best
Sourcing: scouring for new potential customers and customer types
Software Engineering
Avoid going to StackOverflow ever again
Debug and iterate code
Presenting
Text >< images/tables
Data formatting
Speaking and Learning
Querying history for analogies (and next steps)
Super powerful translation and understanding connotations/nuances
Example conversation generation
Deciphering jargon
Exploring different points of view
Fun
Recommendations for books, movies, etc.
On a final note, as we’ve looked for use cases in all parts of the webbing one AI seems to trump the others in terms of personal use: PerplexityAI not least as you can check sources. Useful.
Tikto
At Tikto, we purchase stakes in EBITDA profitable businesses or tech businesses with challenged business models and follow that up with incremental growth capital. We bring our network of experienced operators to help execute a business plan for the next phase of our portfolio companies’ growth.
Get in touch: hello@tiktocapital.com