Playback speed
×
Share post
Share post at current time
0:00
/
0:00

Are humans more deterministic than a Large Langue Model?

𝗜 𝘄𝗼𝘂𝗹𝗱 𝗮𝗿𝗴𝘂𝗲 𝘁𝗵𝗮𝘁 𝘁𝗵𝗲𝘆 𝗮𝗿𝗲 𝗻𝗼𝘁.

I've used large language models to bring a great deal of precision and consistency to the qualitative phase of #JTBD research. While a slightly different 𝗷𝗼𝗯 𝗺𝗮𝗽, or 𝗺𝗲𝘁𝗿𝗶𝗰-𝘀𝗲𝘁 will be produced when running the same inputs multiple times, they are basically consistent in scope. Just slight variations in language.

Humans, on the other hand, introduce a great deal of variability via bias, and are heavily influenced by factors that they don't bother to document for later analysis.

𝗜𝘀 𝘁𝗵𝗶𝘀 𝗮 𝗽𝗿𝗼𝗯𝗹𝗲𝗺?

Directionally, probably not. But, if you 𝘀𝗲𝗲𝗸 𝗽𝗿𝗲𝗰𝗶𝘀𝗶𝗼𝗻 in your research and also demand an audit trail of the thought process, there's nothing better than an LLM at this point in time - as long as you preserve your prompts and inputs.

✅ It's orders of magnitude faster

✅ It's orders of magnitude more consistent

✅ It's orders of magnitude more scalable

✅ It's orders of magnitude less expensive

✅ It's orders of magnitude less invasive of your daily routine

As a result, you can pivot your thinking and scoping before talking to anyone in the field. How many times has a stakeholder asked you to pivot after qualitative interviews, or worse, after fielding a survey?

Practical Innovation w/ Jobs-to-be-Done is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

No, LLM outputs aren't deterministic like the outputs from code is expected to be, but humans were far less deterministic when executing tasks before we had software or assembly lines. No two people did things exactly the same way. There was always nuance - we called this craftsmanship 😉.

Nothing's perfect. No one is perfect. Striving for perfection misses the point and potentially wastes resources by overserving an audience that wants to get the job done better, with less friction, cost, and time.

Are you looking for a differentiated approach to market research? One that has more predictable outcomes and can now be executed faster, and cheaper?


If you'd like #AI accelerators that will set you apart - even if you don't use the formal #JTBD approach - you can get them here

How to plan your JTBD research

How to select the right market to study

Educate yourself and your colleagues with the JTBD Strategy Stack - https://community.zeropivot.us/courses/offers/adc27a2e-9e72-4404-97f2-112cc089f437

Fast track your Journey Analysis with this toolkit

Or, feel free to schedule some time to discuss how my partners and I can help you

~MikeB

Practical Innovation w/ Jobs-to-be-Done
Practical Innovation w/ Jobs-to-be-Done
Mike Boysen shares insights into the evolution of Jobs-to-be-Done, especially in the age of Generative AI. He makes the previously secret process more accessible new approaches and automated tools that vastly reduce the time, effort, and cost of doing what the large enterprises have been investing in for years. This will be especially interesting for the earlier stage, smaller enterprises, and those investing in them who have always had to rely on a superstar, or guess (or maybe that's the same thing!). So...check it out!