AI as an Interface

Have you heard about AIaaI before? Probably not, because I made it up. But I believe this is the future of our digital interaction. Or at least how I imagine it to be.

AI as an Interface will be the omnichannel (single interface) for interactions between any user and any system, application, or service.

“Please order my lunch to be delivered at 2:00 pm to my office. I’d love to have a Caesar salad today, from the same place I ordered from last week. Oh, and please schedule my laundry pickup for 8:00 pm and my car maintenance next Saturday.”

The above scenario isn’t just a far-fetched imagination of a futuristic person like me, this might soon be the mainstream method in our life. AI will become our primary interface for interacting with any service or application.

Right now, we have countless apps on our phones, many of which we’ve used only once but keep around anyway. Imagine if, instead of using those apps, you are only using AI as the only interface for all these apps. This would be an ideal use case for dealing with the exploding number of apps we don't really need.

The internet has transformed our lives in less than 30 years, and most of our jobs now rely on it. Can you imagine working without the internet today? It's gonna be hard, right? The same will soon be true for AI as an interface. It will be the only interface you need to manage your work, apps, and software. These will merely be databases, backends, and system integrations hidden behind the only interface that you really need.

The current wave of AI is still focused on better content creation, be it text, documents, plans, photos, designs, or videos. The upcoming wave will be AI agents that can perform autonomous tasks independently or act as personal assistants. But that's not all. In the near future, we will see AI as the interface for every digital interaction imaginable.

Everyday Use Cases of AI as an Interface

To understand the concept better, consider these use cases, all of which could be just months away:

• “Please schedule my ride to the client for the 11:00 am meeting. Select the best price, but make sure the service provider has good reviews.” Would you care if it's Uber, Lyft, or Careem? Maybe you do, but many people won't.

• “I want to buy a new Bluetooth computer mouse. Check online reviews and prices and get me the best deal. Order it directly to my home.”

• “Please send flowers to my mother to be delivered by 7:00 pm, with a maximum budget of $40. She loves purple color, so make sure the bouquet includes some.”

• “Remind me of my next dentist appointment. If it will be soon, please call them and make sure my dentist is available and not on vacation."

• "Check if any of my medications are running low and need a new prescription. Let me know if I have any blood tests that should be scheduled soon, and remind me to fast 12 hours before.”

• “Plan my family trip for Christmas in London. Send me three suggested itineraries and book hotels with free cancellation. Remind me to make a final decision by Thursday evening so you can book me the best flights and confirm the hotels.”

• “Check the CCTV camera at home and tell me how many times the neighbor’s dog tried to dig in my backyard last week. If there’s a pattern, activate my sprinkler system during those times.”

• “Analyze my sleep cycle for the last two weeks and recommend a new bedtime wind-down approach. The last one was effective, but I need better quality sleep.”

AI in Work Environments

How about cases that interact with your work systems?

• “Send me a report of our monthly transactions for the top five best-selling online products from last year. Analyze whether there are any correlations between age groups, locations, and seasonal trends, and summarize the key insights.”

• “When is our next cash flow low point predicted to be this year? Also, provide specific recommendations to effectively manage and mitigate any potential issues during this period.”

• “Check the action items from our last board meeting. Review if I’ve missed addressing any points in my draft presentation in my personal documents folder, and highlight in yellow color the ones that need to be addressed in a new slide.”

• “Check the LinkedIn profile of "Nader Torki," the one working in AI strategy consultancy. If there are multiple profiles, he is the one with salt-and-pepper hair and a nice beard. Extract his information, add it to our CRM, and let me know if there are opportunities or leads we can work on together.”

• “Verify that our latest planned product upgrades align with the latest regulatory requirements and suggest any compliance initiatives to discuss in our next weekly meeting.”

Making the Vision a Reality

I believe you now get the idea of what AI as an interface could be. Don’t you think this is the way to go? These instructions to the AI interface could be sent as text notes, voice commands, or even WhatsApp messages. They wouldn’t rely on a specific app or a specific service, like Siri, Copilot, Alexa, or ChatGPT.

But these are just my imaginations for now. To make this vision a reality, challenges need to be addressed first. So what needs to be done to make this futuristic, science-fiction vision come true?

1. Address Privacy Concerns

It won’t be easy for people to hand over personal data or confidential organizational information. Today, I still don’t want any system to know my mom’s address or the results of our Q2 board meeting. However, I’d be willing to share that data if I were assured that it was secure, private, encrypted, and accessible only by me. Current cloud providers (without mentioning any names) claim to offer these guarantees, but we still have our privacy doubts and concerns, as it remains centralized in the hands of a few private sector companies, limiting trust and control over personal data. Maybe more regulations need to be in place. Maybe a new concept of “Personal Cloud” needs to emerge. Maybe the future is in the decentralized blockchain? A lot of possibilities and questions need to be addressed before allowing our precious data to be shared.

2. Apps Should Allow External Access Over Their Services

Software and app developers are still operating as isolated ecosystems with strong defenses and secure gates. Apps and software rarely communicate, especially if developed by different companies. The integration between systems or apps currently happens mostly at the back-end level rather than seamlessly at the user interface level. This needs to change, and I’m not talking about APIs here with specific prerequisite definitions. We need a more seamless form of external access where different applications and services can interact freely without strict protocols. Today, I still can’t just tell Siri or ChatGPT to order lunch from Deliveroo or order my coffee beans from Amazon. They don’t allow it now, but a disruption in this process needs to take place.

3. Better Digital Identity Management

For the above two points to work, we need better digital identity management. Human identities should be verified by biometrics or other means alongside a unique system identity. This means my AI interface identity should also be verified, and any service I use, like Uber, should have a verified unique identity on my phone or computer, and it should only be able to handshake my Amazon account unique digital identity for them to work together. 

4. Data Analysis 2.0

AI thrives on data, but we still don’t have quality of data over quantity of it. Machine learning and data science are still focused on structuring big data, but I believe this is about to change. You don’t need to go to this structure level anymore, thanks to the neural networks in place now. Data analysis 2.0 should be able to effectively work with raw data without structure and make sense of it in real-time. Technology is there; we just need to tweak the architecture a little bit to reach that point.

5. Infrastructure and Connectivity

The above perspectives will be hungry for processing power and fast bandwidths that should be available on each and every device we use, and not only on the cloud. There is a lot of room for improvement here beyond just the limited 5G coverage, expensive GPUs, and centralized data centers. We can start with what we have, but a new wave of distributions and innovative ideas in this field needs to be discussed broadly.

---

Shaping the Future

The future is not only something that comes knocking on your door with some surprises. It’s something that we need to imagine, plan for, and start exploring its possibilities from today to shape it gradually.