AI at the System Level

#AI #Middleware

The opinions in this article are my own and do not necessarily reflect the views of my employer.

Laura is gently awakened from her sleep by a soft, familiar voice. It's Athena, her AI Assistant, sharing a summary of the day's weather and Laura’s upcoming meetings. Laura steps out of bed and heads to the small gym in her apartment, guided by Athena's morning workout suggestions based on her sleep pattern and vital signs.

After her workout, Laura steps in to her smart closet, where Athena, having connected with her calendar, suggests a professional ensemble perfect for a mid-day executive meeting. As Laura heads downstairs to her kitchen, her coffee machine, already brewing her favorite blend, cheerfully reminds her that it needs descaling.

As Laura sips her coffee, Athena filters through work messages, separating the important ones from Laura’s colleagues and clients and summarizing the less urgent ones. Through Laura's employer’s enterprise middleware, Athena integrates with the AIs for the CRM and finance systems. Working with these AIs, Athena analyzes Laura’s sales pipeline and prioritizes the high-margin deals that can potentially be closed before the end of the quarter. Athena discusses the analysis with Laura, and together, they decide where to focus to maximize the sales numbers.

Laura then moves to her home office, where her first task is to draft a statement of work for a client. Athena seamlessly integrates with their firm's AIs for the legal and document management systems, pulling up the necessary templates and directory of approved legal terminology to draft an impeccable SOW customized to Laura's usual writing style and company standards. Athena then puts the document into Laura's workflow for review and approval.

As the day’s activities accelerate, Athena leverages enterprise and public systems (all fronted by AIs) to analyze Laura’s emails, phone calls, video calls, documents, tasks, and more and then prioritize and pre-process them. Working as a team, Athena and Laura “get into the zone,” and they have a productive and fulfilling workday.

In the late afternoon, Athena reminds Laura about her dinner plans. As Laura wraps up her work, Athena helps her switch gears. She integrates with Laura’s health service to suggest relaxation techniques based on Laura’s biometric data and then integrates with Laura’s entertainment service to cue up the optimal playlist. When Laura heads out to meet her date for dinner, Athena touches base with the restaurant’s AI, and they review Laura's taste preferences and dietary restrictions to select the perfect dish and drink pairings. Athena also prompts Laura with recent personal developments of her date, gleaned from an information exchange with her date’s AI Assistant, to ensure a more personalized and warm conversation.

As she returns home, Laura gives a mental nod of thanks to Athena, her ever-present AI companion. Athena wishes her a good night and begins running systems checks around the apartment, ensuring a quiet and restful environment for Laura. Athena processes the day's data, preparing to make tomorrow another efficient day in Laura's life. In their highly interconnected world, AI Assistants like Athena are not just tools but personalized companions, making every aspect of life simpler, more efficient, and highly customized.

How much of this is available today?

• AI Assistants: Digital assistants like Google Assistant, Siri, and Alexa can perform tasks such as setting reminders, answering questions, controlling smart home devices, sending messages, and providing news and weather updates.
• Smart Home Devices: Smart thermostats, smart lights, and smart locks are commonly used today. They can be controlled via digital assistants, mobile apps, or even autonomously operate based on learned patterns.
• Fitness Recommendations: Fitness and health apps can suggest workouts based on a user's goals, activity levels, and health data.
• Email Filtering: AI can help filter and categorize emails with services like Google's Priority Inbox or Microsoft's Focused Inbox.

How much of this will be available soon (within a couple of years):

• Advanced AI Integration: While AI can perform tasks individually, the seamless integration of AI Assistants across multiple domains (like legal, sales, and personal lifestyle management) is still a work in progress.
• Personalized AI Interactions: As AI models continue to improve, we will see more personalized and human-like interactions with AI Assistants, including understanding user moods, nuances, and preferences in greater depth.

How much of this is more than a couple of years away:

• Full Home Automation: The scenario where all devices in a home, including items like the coffee machine and closet, are 'smart' and AI-integrated is still a few years away from being commonplace. While smart appliances exist, the level of integration described will require significant advances in IoT and AI technologies.
• Advanced Biometric Monitoring: While there are devices that can monitor heart rate, sleep patterns, etc., the level of integration and proactive response by an AI (like suggesting relaxation techniques based on biometric data) is currently in the nascent stages.
• AI Powered Business Decisions: The concept of an AI Assistant capable of making strategic business decisions like identifying potential deals in a CRM is under development but is still several years away from being commonplace.

In the fully AI-enabled world, AI-fronted systems won’t just be reactive but will also be proactive, capable of anticipating needs, interacting with each other, and even generating innovative ideas. This will create a complex and highly adaptive IT landscape that will bring new and exciting design and development challenges.

The UX/UI of this landscape will be the AI Assistants for individuals, serving as gatekeepers and personal managers. They will negotiate, integrate, and prioritize information from various public and business system AIs – CRM, Finance, HR, middleware, and more – to create a comprehensive, streamlined experience for their human counterpart.

Business system AIs will respond to inquiries and actively engage with each other, middleware, and the AI Assistants to share data and communicate vital information and insights.

AI middleware will play a pivotal role in this enterprise architecture. The middleware will surface fresh, original insights by consuming, analyzing, transforming, and extending information from multiple enterprise and public sources. Middleware will be the "innovation hub" of the IT landscape, where disparate information streams converge, and novel ideas emerge. Instead of just moving data around, the middleware will both orchestrate the flow of, and create, information, knowledge, and wisdom.

This complex interplay of AI systems will fundamentally change IT development. Development will shift from creating individual, standalone systems to engineering an intricate, interconnected web of AIs, each with its own information, roles, and responsibilities. This will necessitate a new approach to system design – one that prioritizes seamless integration, autonomous operation, and constant communication between systems.

Moreover, with AI systems constantly learning and self-improving, development and operations will genuinely merge and become a continuous process. It will no longer be about deploying a release and moving on to the next development phase; it will be about nurturing, refining, and adapting these AI entities in real-time, ensuring they function optimally in the ever-evolving digital ecosystem. The AI entities will not just be fault tolerant but will also self-repair and self-update. This will necessitate development teams having a deep understanding of human behavior, AI behavior, tactical and strategic business needs, as well as the more traditional technology skills of today.

In this future landscape, enterprise IT is more than just infrastructure – it becomes a thriving, evolving digital organism, with AI systems as its constituent cells, each playing its part in maintaining health and advancing the evolution of the whole. This is equally exciting and challenging, heralding a new era of “interesting times” in IT system development and operations.