You don ever think about di difference between tool and system?
Tools na wetin we dey use do work. Systems too dey make work fast and easy.
Some pipo fit think say system na just tool wey don too complex.
But, if dem divide work into two types—iteration work and flow work—di difference between tools and systems go come out clear.
Iteration and Flow
Iteration work na di way dem dey take build wetin dem want small small, by trying different things and correcting mistakes as dem dey go.
For iteration work, one set of tools wey dem fit use for different specific tasks dey helpful well well.
But, flow work na process wey dey move step-by-step, and e dey produce wetin dem want for di last stage.
For flow work, if you get system wey dey guide di flow of work, e go make productivity and quality better well well.
Flow Work Conversion and Systematization
Many works wey humans dey do na either iteration work or part of flow work wey dem don arrange well.
If dem change iteration work to flow work and then arrange am as a system, e go make productivity and quality better well well.
Di Industrial Revolution ati Di IT Revolution
Di Industrial Revolution ati di IT Revolution na good examples of how changing iteration work to flow work, and then making am into a system, make productivity and quality better well well.
Before di Industrial Revolution, dem dey do manufacturing as iteration work. Humans dey use tools skillfully and dey free to change how dem arrange things and di steps dem dey take each time.
Na so too, before di IT Revolution, processing information mean say humans dey use tools for one unorganized, iteration way.
By making these processes into systems, just like factory production lines or business IT systems, dem make productivity and quality better well well.
But, no be only making am into system, but also di flow work conversion of iteration tasks na very very important. Na because dem fit change am to flow work first make am possible to make am into a system.
Di Generative AI Revolution
When we dey try to make productivity and quality better by using generative AI for business, just using AI as a tool no go bring out its full value.
The main aim na to change iteration work to flow work, and then to make that flow work into a system.
Generative AI fit handle iteration work because e fit adapt well. But, whether na humans or generative AI do am, iteration work get limits to how productive and good e fit be.
Because of this, trying to change to flow work and making it a system na very important.
Some pipo fit argue say if flow work conversion fit make productivity and quality better for human workers, dem for don do am even before generative AI come out.
However, flow work conversion, when e depend on human workers, na actually a very hard problem. Human workers no fit quickly adapt to changes for wetin dem suppose do or di content of di work.
But, if di worker na generative AI, e easy to change roles and task content plenty times.
Unlike humans, generative AI fit forget previous steps, quickly read and understand new procedures, and do tasks based on dem.
For this reason, di main way to use generative AI for business na to change iteration work to flow work and then make am a system.
How Generative AI Dey Make Business Easy
Make we look at one example of how generative AI dey make business easy.
For example, consider di work of answering employee questions about company internal rules.
Dem fit use generative AI to search internal rules and prepare answer drafts.
However, e fit happen say generative AI go refer to old rules or mistakenly create answers based on imagined information wey no dey for di rules.
Also, questions fit come through different ways, like email, messenger apps, phone, or in person.
So, di employee wey dey handle questions still need to receive dem as before.
E make sense say dem fit make am more efficient if employees answer questions straight away if dem fit, and for questions wey need rule verification, dem go put di question inside generative AI to generate draft answers.
On top of dat, for questions wey dem dey ask often, e necessary to post dem on di company's internal website as FAQs.
Generative AI fit also be used to put common questions and answers and generate bulleted drafts to publish on di website.
Wetin pass dat, dem fit use generative AI to review draft wording when dem need to amend rules.
Dis kind uses fit make some percentage of question handling tasks easy.
But, dis one na just treating question handling as iteration work and using generative AI as a tool.
As a result, di efficiency gains from dis approach na very small.
Flow Work Conversion
To make di inquiry response work wey we mention as example reach maximum efficiency, we must change dis work to flow work.
Dis one need us to go deep and write down di steps wey di person in charge dey take when e dey handle inquiries:
- Receive inquiries through different different channels.
- If di inquiry na di same with one wey dem don answer before and no change for di rules wey concern am, give di same answer as before.
- For new inquiry, or one wey concern rule change, check di rules and prepare draft answer.
- Check if di draft answer refer to old rules or e include information wey dem no state for di rules.
- Check if dem need approval before answering, and get approval if e necessary.
- Answer through di channel wey dem receive di inquiry from.
- Register di inquiry content, approval result, and answer result for di inquiry history data.
- Dey check di inquiry history data regularly to create suggested updates for frequently asked questions and answers.
- Update di company's internal website after getting approval.
- When dem update rules, update di rule data wey dem refer to.
- At di same time, record for past inquiry history data say related answers and rule updates don happen.
- Verify if frequently asked questions and answers need revision because of rule changes, and update if e necessary.
By making di details of these tasks clear, and connecting dem, di flexible iteration work fit change to a clear flow work.
Example of Systematization
By changing works to flow work, di way to make am into system go clear.
When you dey make am into system, if e no too much say employee no too convenient, one way na to gather all di inquiry channels to one place.
But, if employee convenience na wetin dem value pass, all di inquiry channels suppose still dey open.
Basically, di system suppose collect inquiries directly. Only if na verbal inquiry, human go put di details into di system.
After dem collect one inquiry, di IT system and generative AI go do as much of di next works as possible according to di flow. At di beginning, human go dey check and approve small small throughout di system, and human operators suppose fit correct things.
Then, as dem dey use di system to handle inquiries, if di generative AI make mistake, dem suppose update di instructions for di AI with things to watch out for, things to check, examples of mistakes, and correct examples to stop di mistake from happening again.
Dis process fit reduce generative AI's mistakes. Changing how dem dey update these AI instructions from iteration work to flow work fit make am even more efficient.
For dis way, by making flow-converted tasks into a system, even works wey for first look like say na human suppose do am fit be replaced by a system wey generative AI dey lead.
Common Misconceptions
Plenty pipo believe say using generative AI for business now no really get much effect, or say e still too early.
However, most of these pipo often get two kyn misconceptions.
Di first misconception come from say dem dey focus on using generative AI as just a tool.
As we don show here, using generative AI as a tool for iteration work no really make business efficiency better. Dis misconception come from experiencing or seeing such small results.
Di second misconception come from say dem dey focus on making generative AI do iteration work.
True true, trying to make current generative AI do iteration work often no dey work. So, pipo dey mistakenly conclude say generative AI no fit take over tasks wey humans dey do, just based on dis observation alone.
Conclusion
As we don discuss, if you change iteration work to flow work and make am a system, you go get more efficiency pass just using tools alone.
Wetin pass dat, even if generative AI no fit do iteration work, e fit handle plenty individual tasks inside a flow work process. Even if mistakes dey plenty at first, you fit still dey improve am steady by updating di instructions.
Alternatively, if e necessary, tasks fit be divided, separating writing drafts from checking, or implementing checks for many stages.
If dem fit achieve systematization for dis way, then improvements go dey happen with each task dem do, and operations go come dey more efficient as time dey go.
Dis na one way of working wey dey allow di mechanism itself to dey improve continuously, just like factory production and IT systemization.
To effectively use generative AI, you need to change your mindset: instead of trying to improve your own iteration work, you must objectively change your tasks into flow work and make dem a system.