Artificial Intelligence, Machine learning, Expert systems, and Neural networks have evolved over time and recently gained tremendous popularity after the release of large language model (LLM) based AI i.e., ChatGPT. AI products and tools are helping humans increase their efficiency by generating content, creating custom images, exploring of ideas, and much more.
On a lighter note, here is a picture which shows the penetration and impact of AI products in our society through attempt, acceptance, and adaptation.
For example, Designers can use AI tools for the following:
Quickly summarise and synthesis the user research (conducted and documented by them).
To pull out meaningful insights and perform sentiment analysis, and by uploading data (like surveys, customer/employee management systems, live-chats, social media, and more) to AI tools.
Below are some AI tools that can augment the UX process and creativity:
With AI also comes a responsibility to use it in the right manner to avoid any unexpected outcome. To understand this in a better fashion we will cover:
AI Pitfalls: Areas which every user should be aware of to avoid risks.
Coforge’s perspective: Recommendations to use AI for consulting assignments with sensitive data and ensure security for our clients.
We must be incredibly careful while working with open-source AI, especially with confidential information because we are not sure how data will be stored and processed further.
For example: This is the message we get when we start using ChatGPT
Below are a few recommendations to avoid the pitfalls:
Garbage In, Garbage Out
Always feed right and ethical data (responsibility of data scientists) to AI or it could lead to unexpected outcomes. AI should be trained on a variety of data to avoid stereotyping
We should ask when evaluating a dataset:
Where did the dataset come from?
What was the method of data collection?
If it was survey data, what are the assumptions and conditions under which this data was obtained?
Were any of the data imputed (missing cells filled algorithmically)?
What other datasets could be joined to add supplemental context?
What do subject matter experts know about the data and how could this knowledge be beneficial to learning?
Safety and Security
Always prompt ethically in the correct manner. If someone intentionally tries to exploit the AI then that is a safety concern and we do not know how to fix that. For example, someone tried to exploit Bing and it became offensive and was taken down.
Trust and transparency
We do not know the result given by AI is reliable or bogus because AI systems are not transparent to us, and we need so much knowledge to understand what is going on behind. To gain trust AI systems should:
Add the sources
Show the sources
Tell the steps that AI took to get that answer
Know where AI is getting its results from and let users know
Ownership and Intellectual property
AI systems do not have ownership or Intellectual property rights because if an AI system recreates Monalisa than it will not have same value as DA Vinci’s Monalisa because it has a narrative behind it and it is human. We value human things.
Patent law considers the inventor as the first owner of the invention. The inventor is the person who creates the invention. In the case of autonomous AI generating an invention, there is no legal owner as the AI technology cannot own the invention.
Do not over rely on AI
Overreliance on AI leads to:
Lack of creativity and innovation: It is important for designers to work closely with end users, stakeholders and use user-centred design kit.
Illusion of objectivity: This means using technology to objectify is not true always. For example, how often do we watch something suggested by NETFLIX
Coforge designers and product development teams work on sensitive consulting assignments and deal with confidential client data. AI and contextual experiences are driven by data and recommendations. And dealing with any sort of customer data means it is vital for the longevity of our investments.
So, for Coforge, privacy, governance and ethics come first.
Working with AI gives us opportunities to:
Create hyper personalized experiences
Speed up the design process through analysis of user research and data
Generate proto persona, scenarios, user flows, content, color palette and suggestions for the initial brainstorming
Explore creative freedom by seeking suggestions and ideas
Here are our recommendations to avoid any unexpected outcome while working with AI:
Machine learning is based on recommendations, and it must be done in a way that respects our customers’ information. So, we need to keep privacy on top of our customer experience design mandate.
Always question how to deliver personalized experiences in ways that are localized and aren’t reliant on third-party sharing. For example: Is it safe to share user research data with AI for analysis?
Leaders should become more contextual and accountable in their data. They should draw clear ownership and responsibility lines throughout AI frameworks to sustain trust across all stakeholders and it should go under consistent review and monitoring.
Prioritize ethical considerations and understand the risk surrounding the unintended consequences.
Built technology by keeping human society and the environment in mind. It should be built with human-centred values at the core and must adhere to privacy, security, and protection laws.
Machines can generate ideas, but they don’t have senses. Even the best of guesses needs expert validation. That’s where Coforge’s designers and consultants add value by leveraging the AI ethically and by taking creative decisions on the basis of critical thinking and domain expertise.