AI: not only in the rational realm
A black box effect
AI delivers impressive results while remaining mysterious. Bold communications amplify this perception.
Commercial promises and alarmism
Communications from AI players oscillate between incredible promises and sometimes conspiratorial alarmism.
A runaway public opinion?
Hard to stay calm. Two observations:
- Google Duplex, announced in 2018, is now seeing concrete applications come to life.
- Some announce the « end of creative humanity », starting with illustrators. If an incredible announcement from 2018 is coming true, why not the other?
Do we need to take a stand?
Jobs will be created, others will evolve. Between fears and hopes, everything is moving « fast and hard ». Spoiler: we are working on it and remain confident about humanity.
On a personal level
The author programmed his first video game in 1981. ChatGPT was a shock. AI will DISRUPT some domains (taking no prisoners) while others will remain virtually untouched.
A creative technical solution that can lead to innovation is not a « statistical average » of big data, unlike a face generated by AI.
A definition of creativity
Originality AND value
To be creative, a result must meet two conditions:
- It must be original: something good but already seen is not creative.
- It must bring something: new but useless is just new, not creative.
We propose that, to be considered creative, a result should be both original and bring something of value.
The illusion of creativity
Regenerative AI gives a sense of originality through its recombinations. You can re-prompt until you get a result you like: it seems original + interesting = Bingo. This works for images, cinema, text... but what about technical domains that demand precision?
Requirements vary by domain
A drawing with 6 fingers? We can live with that. But a phone number generated by combining several doctors' numbers is useless.
Combining technical solutions rarely produces an original solution that actually works.
A creative result that is precise and functional
Combining images and text can yield good results, but it is very risky for technical solutions.
There are very few technical solutions that work, unlike the vast number of possible graphic illustrations, because the latter do not have to comply with as many physical laws.
Regenerative AI & / Vs creativity
Predictability by definition
If AI can generate THE BEST new idea on demand, it becomes by definition predictable, therefore anti-creative.
The myth of the best idea
If AI finds the ultimate idea (like 42 in Douglas Adams), it will be found only once. But there is no lasting ideal solution. The best solution depends on many parameters (see TRIZ).
Using AI for quantity
AI can flood us with millions of ideas, but what is their value? You might as well give a keyboard to a monkey to write a patent.
The role of AI
AI will not replace the ideation phase
Predictable and imprecise, AI does not suggest a capacity to generate ideas that are both creative AND technically precise.
Acting differently, elsewhere?
AI can help upstream, by supporting proven methods such as TRIZ or ASIT.
SolidCreativity's proposals
After months of development, we are opening the beta test of two online tools: TRIZ matrix optimization powered by AI, and an ASIT method facilitation tool.
*regenerative: our proposed naming
See also: You « innovate » with AI? LOL! and AI Vs Brainstorming: what if innovation were elsewhere?
