A new era for UX, thanks to AI: an explainer for product professionals on tomorrow's standards

Сreative or generative AI

Generative AI has made significant leaps forward and its influence is far-reaching. Industries such as healthcare, retail, and manufacturing are employing generative AI to predict future scenarios, optimize processes, and customize products. An IDC report predicts that by 2025, at least 50% of large enterprises will use generative AI for planning and scenario analyses.

For instance, in healthcare, generative AI can help generate digital models of human organs for better diagnoses and surgical planning. In retail, it is used to personalize product designs, catering to each consumer.

The latest breakthroughs in image generation

Stable Diffusion XL offers a transformative experience for non-artists. The rapidly growing open-source AI platform empowers non-artist to create descriptive 3D images with shorter prompts. The platform boasts next-level photorealism capabilities and enhances image composition and face generation.

Adobe firefly offers a transformative experience for non-artists The platform provides “virtual models” for hire, which can be used in diverse settings, from social media content to e-commerce product photography.

Deep Agency, an AI-powered virtual photo studio and modeling agency, is at the forefront of a rapidly developing and controversial industry. The platform provides “virtual models” for hire, which can be used in diverse settings, from social media content to e-commerce product photography.

AI features that will revolutionize online meeting tools | Cieden

Photo / document scanning
Using AI for onboarding, for example, will allow users to scan ID/documents and automatically populate relevant fields, leading to swift, accurate entries. This capability is particularly beneficial for banking apps, travel platforms, and identity verification services.

Fill in form from unstructured input & errors autocorrect
Integrating AI for form management can streamline user experience. For instance, AI can process unstructured inputs, intelligently sorting data into appropriate fields. It can also detect inconsistencies or errors, such as mismatched zip codes and city names, and suggest accurate alternatives. Such advancements can be invaluable for e-commerce sites, real estate platforms, and registration systems.

The future of user interaction with digital products

LMM superpower enables new functionlaity

As we mentioned before, AI now stands at the forefront of user experience, shaping how users interact, engage, and derive value. AI can act now as an advanced assistant, not just pulling information but actively molding experiences. It's exceptional at analyzing and interpreting data, summarizing vast amounts of information, and even tailoring content styles to fit specific user needs. Classification is one of AI's standout capabilities. It's adept at sentiment analysis, topic detection, fraud recognition, and much more. Such nuanced classifications can be crucial in areas like risk assessment, enhancing customer service, or even boosting user engagement.

On the reasoning front, AI brings efficiency to the table. From managing tasks and setting deadlines to suggesting optimal data presentations, AI streamlines many processes that traditionally demanded considerable time and effort. It even goes a step further, self-assessing its outputs, adapting to user preferences, and navigating intricate workflows without a hitch

However, with all these advancements, it's vital to stay aware of the limitations, especially if you're considering integrating AI into your products or launching an AI-based startup. While AI models like ChatGPT have revolutionized text generation and interaction, they aren't flawless. They can sometimes produce inaccurate or even 'hallucinated' outputs. Their understanding of context, especially in real-time, remains a challenge. Moreover, integrating them seamlessly into various platforms isn't always straightforward. There's also a matter of limited multimodal interaction capabilities, potential data privacy concerns, and issues with enterprise-specific adaptability. As powerful as they are, current AI systems still have room to grow and improve.

In this chapter, we'll delve deeper into the potential of AI-enhanced user interactions, drawing insights from both emerging startups and established companies that have recently unveiled innovative AI features. 

FlutterFlow AI Gen offered innovation in visual app building

FlutterFlow is built by two former Google engineers who wanted to simplify the journey of creating an app and make it more accessible to designers, developers, and entrepreneurs. They aimed for a simple drag-and-drop interface that lets them build a fully functioning app within an hour. With just a simple prompt, the tool generates both visually captivating app designs and the necessary code, which can be imported into FlutterFlow with ease.

This all runs on the ChatGPT API, offering a user-friendly experience across both mobile and Mac OS platforms. The tool also instantly turns textual prompts into project structures, making it easier to customise designs or export generated code.

Additionally, it quickly creates backend schema, transforming your ideas into scalable and interconnected databases. One of the standout features is an AI-driven color scheme generator that crafts unique and aesthetically pleasing color palettes based on text descriptions. Within the custom code editor, users can benefit from real-time AI-powered code suggestions, further enhancing the app development process.

Buzzy announced Figma plugin that allows designers to transform text into fully functional applications without coding knowledge

As stated, Buzzy operates directly within the Figma design tool, enabling users to turn their app or website designs into a functional prototype or even a live application.

Here's how it may work:

1. Design. You begin by creating your user interface (UI) design in Figma. This can include anything from the layout of your app, the placement of buttons, forms, text fields, and so on. You can also import pre-existing Figma designs.
2. Annotate. Once your design is ready, you add annotations using Buzzy. These annotations are instructions that help Buzzy understand what each element of your design is supposed to do. For example, you might annotate a button to perform a certain action when clicked, or a text field to display data from a certain source.
3. Transform. After you've annotated your design, Buzzy's AI engine gets to work. It interprets your annotations and transforms your design into a working app. This isn't just a static prototype - it's a real, live app that you can interact with, test, and refine.
4. Deploy. Once you're happy with your app, you can deploy it directly from Buzzy. There's no need to worry about servers, hosting, or scaling - Buzzy takes care of all that for you.

Since its inception in 2016, Brain AI has been a force in the realm of natural language processing. In 2021, Brain AI furthered its legacy with "Natural AI", a digital commerce solution flaunting the world's first generative interface, refining how we engage with AI.

Come March 29th, Brain AI introduced its newest gem: Imagica. This cutting-edge, no-code AI toolkit offers users an intuitive platform to breathe life into their concepts, regardless of their tech know-how. It boasts real-time data accessibility, versatile functions, and rapid performance, positioning Imagica as a go-to hub for swift app development.

While Imagica's exact nature - be it a trailblazing app or a software-as-a-service offering - remains under wraps with its beta-phase status, its premise of "empowering users to shape AIs that change the world" is undeniably compelling. Users can weave text, visuals, clips, and 3D models into their app visions. Powering this is Natural OS, offering a rich palette of 4 million functions for app crafting.

Collaborative robots for repeated or exhausting tasks

Market analysts predict a surge in companies employing AI-equipped machinery to execute repetitive and physically deman-ding tasks. This shift allows human personnel to focus on more specialized roles. AI capabilities can also empower teams to rapidly identify and address faults or malfunctions, enhancing security and minimizing expenses related to repairs or injuries. Cobots are anticipated to become increasingly common in sectors such as:

• Healthcare and hospitality. Tasks include sample gathering, replenishment of hospital supplies, surgical assistance, recovery from injury, and support for healthcare workers in homes for the elderly or people with disabilities.
• Food. Warehousing, food packaging.
• Agriculture. Usage of drones for tasks like seed sowing, fertilizer and pesticide dispersal, tracking intruders and invasive species, and implementing LED lighting and hydroponics in indoor farming environments.
• Electronics. Quality inspection for smartphone chips, chip processors, and printed circuit boards.
• Emerging tech. Incorporation of torque sensors, proximity detection sensors, end-effectors (like vacuum, mechanical, pneumatic, and magnetic grippers at the end of robotic arms).
• Defense. Road clearance of explosive devices, explosive-detection sensors.

AI assistence extends to navigating the internet 

Innovative startup HyperWrite recently revealed an AI agent capable of navigating and interacting with websites. CEO Matt Shumer demonstrated the AI's abilities, such as completing an online order on Domino's Pizza, via their Chrome extension. Distinguishing itself by learning from past interactions, HyperWrite's AI promises to usher in a new era of web automation and personalized assistance. Though promising in automating web tasks, concerns around security like phishing, hacking, and fraud were addressed, with Shumer assuring vigilance on these issues. So, there is still a lot of work ahead in this regard.

In the gaming arena, 2023 introduced Voyager, a Minecraft bot. Termed an "embodied agent," it represents a new breed of AI that doesn't just process data but actively explores and interacts with its environment – in this case, the virtual world of Minecraft. It writes its own code using GPT-4, evolving its gameplay strategies on the go.

Minecraft's resemblance to a simplified real-world simulation makes it a prime testing ground for such AI development, with Voyager exemplifying the AI's potential to learn and adapt in dynamic environments.

Its training ground, MineDojo, taps into diverse digital resources, from YouTube to Reddit. The result? An AI that strategizes, learns, and evolves, storing skills and ensuring no repetitive learning - an embodiment of the term "embodied agent.

Dynamic interfaces transform the traditional approach to UX

Generative AI's ability to customize designs at an individual level can lead to a paradigm shift in UX/UI design. It allows for "design personalization", where user interfaces dynamically adapt to each user's needs and preferences, offering a significantly enhanced user experience. This for sure leads to more immersive, interactive, and personalized digital experiences.

This level of customization is a significant step beyond traditional personalization features such as user profiles or personalized content. It allows each interaction with the app to be a unique, tailored experience that caters precisely to the user's needs at that moment. 

In addition, generative AI can facilitate A/B testing of different interface designs in real-time. Based on user responses, the most effective design elements can be identified and incorporated dynamically, leading to an iterative design process that continually improves the user experience.

Personalization becomes individual-driven

The latest breakthrough in personalization, enabled by machine learning, centers around hyper-personalization, a step beyond traditional personalization. 

Instead of segmenting customers into broad groups, hyper-personalization focuses on the individual. 

Leveraging real-time data and advanced machine learning algorithms, businesses can create highly individualized interactions that improve user experience and customer loyalty.

For instance, companies like GlaxoSmithKline, MasterCard, Walmart, Uber, and Sony have developed advanced AI-powered services that can provide personalized experiences based on real-time user interactions and historical data.

Settings (configuration) management advancements

In an age where technology is the cornerstone of business, IT configuration management has become a mission-critical task for companies. However, traditional methods often fall short. They are time-consuming and susceptible to error - a ticking time bomb in a digital landscape that demands efficiency and accuracy. Here's where AI swoops in to save the day. Let's break it down with some real-world examples.

Let’s take Netflix, for instance. They manage a massive global IT infrastructure that streams show to millions of viewers simultaneously. With thousands of servers, managing configurations manually is impractical. Instead, they've employed AI to automatically track and manage these configurations.

The AI system can instantly detect when a server is underperforming or when there's a change in configuration, and automatically adjust settings to ensure optimal streaming quality.

On the other end of the spectrum, let’s consider a small startup like FooBar Inc. They might not have the budget to employ a dedicated IT team to manage their systems. AI can fill this gap. Instead of spending time on manually configuring IT components, FooBar's team can utilize an AI solution like Puppet, which automates the deployment and management of their software across various platforms. This allows the team to focus on their core business, while AI handles the technical details.

Changes in data processing and automation that your product needs to reflect

The rise of agent-oriented programming

The rise of agent-oriented programming (AOP) has been meteoric, especially between 2022 and 2023. For those new to this concept, AOP introduces a software design approach focusing on creating independent agents. These agents, which are autonomous entities, can perceive, react to, and interact with their surroundings to accomplish specific tasks. They can function independently but also have the ability to collaborate, communicating seamlessly with one another to tackle more intricate goals.

Pinecone succinctly described the role of agents by comparing them to auxiliary tools for Large Language Models (LLMs): “just as a human might use a calculator for calculations or initiate a Google search for information, agents act as instruments for LLMs to achieve analogous outcomes”.

One recent breakthrough in AOP involves autonomous vehicle technology. In 2022, RoboCar Inc., an emerging autonomous vehicle startup, leveraged AOP to develop autonomous agents that simulate the behaviours and decision-making processes of human drivers. These agents constantly interact with their environment, making decisions on vehicle speed, direction, and braking based on road conditions, traffic signals, and the presence of other vehicles and pedestrians. 

It's worth noting the advent of the LongChain framework, constructed with LLMs at its core. This framework can be employed across various applications such as chatbots, Generative Question-Answering (GQA), text summarization, and beyond. LongChain's brilliance lies in its ability to "chain" diverse components, facilitating the creation of more intricate use-cases anchored around LLMs.

Furthermore, the LongChain agents possess a remarkable capability. Input a task in textual form, and these agents will craft a comprehensive plan for its execution. Not stopping there, they can initiate the plan, dynamically altering it based on real-time feedback until the desired outcome is achieved. This adaptive planning and execution mechanism is emblematic of the flexibility and intelligence that AOP brings to software development.

AI-powered prompts 

In the 2023 software development landscape, the integration of AI-driven prompts has optimized the way developers interact with underlying algorithms. On platforms like Coda, what previously required extensive coding can now be communicated through succinct prompts. For instance, a prompt like "Group action items by individual and prioritize by deadline" acts as an interface to trigger complex underlying algorithms. The AI model, equipped with its vast training and algorithms, interprets and executes tasks based on these prompts. This evolution has profound implications for the business world, especially startups, as the timeline to develop MVPs has seen a significant reduction.

It is worth mentioning, that recently Coda launched the "AI at Work Challenge”, attracting over 1,000 participants. The outcome? Over 150 templates addressing how AI can aid daily work in areas like research, product development, marketing, and more. Explore the top templates here.

GitHub Copilot, for example, greatly streamlines the coding process with the use of inline commands. For instance, if a developer is working on a Python script and requires a specific function, rather than searching online for the correct syntax and parameters, they can simply type a command in the form of a comment, with no need to switch.

One click instead of text input

AI tools like "Tappy, your AI LinkedIn assistant" have revolutio-nized data processing by enabling dynamic interaction with digital content. Its key feature, the ability to choose the senti-ment of a comment (for example whether arguing, supporting, or questioning) — paves the way for more personalized and context-appropriate interactions, especially for sales prospecting or personal brand development.

Let's take another example of Microsoft 365 Copilot. Imagine that you need to create a proposal. Rather than manually digging through files and manually typing out the proposal. With Microsoft 365 Copilot, all you need to do is instruct it to create a proposal based on the meeting and include product offers from the product roadmap document.
This AI-based automation, coupled with the powerful suite of tools offered by Microsoft 365, represents the future of work.

Text requests as opposed to applying filters: Atlassian

Emphasizing intuitive interactions, they introduced a feature where users can make text requests to fetch information, as opposed to applying filters manually. To ensure a user the applied filters were correct, the system shows the code section.

With natural language processing at its core, this feature understands users' queries, digs into the relevant data, and presents precise results. 

It also employs Machine Learning algorithms to understand team behaviors and project requirements, predicting and recommending the best workflow patterns. Furthermore, the AI can take over routine tasks, such as setting reminders or following up on tasks, freeing up human resources for more complex tasks.

Revolutionizing CRM practices with AI autocompletion

Salesforce's report revealing that 73% of customers anticipate businesses to comprehend their needs without manual input. 

Leading this evolution is the AI-powered autocomplete fields, a transformative technology that harnesses Natural Language Processing (NLP) to extract essential data from text inputs for efficient CRM querying.

Juniper Research predict that by 2025, such AI-driven automation could cut down data entry times by up to 60%, dramatically increasing business productivity. Moreover, Gartner projects a 25% reduction in data entry errors by 2025, courtesy of AI-enhanced predictive coding. 

This surge indicates an imminent shift to more automated, precise, and user-friendly CRM systems by decade-end.

Our experiment showcases AI's transformative potential in CRM practices. Utilizing a sales rep's email, our AI model, ChatGPT, discerned critical deal specifics and automatically populated corresponding CRM fields, demonstrating a significant leap in data entry efficiency and precision.

Removing need of input for commit summary

Imagine that you will never need to write reports on what was done anymore. 

This vision is now closer to reality, thanks to GitHub's innovative AI-powered feature, which autonomously generates commit summaries. The mechanism harnesses NLP algorithms to analyze code changes, understanding the context and essence of each alteration, and then succinctly summarizes it. 

This transformative functionality echoes the strides being made across various industries.

For instance, in healthcare, AI has been employed to read patient data and generate diagnosis reports, with a 2021 study in Nature Medicine showing a 25% reduction in diagnostic errors. Similarly, content creation tools like Jasper.ai have employed AI to automatically draft summaries for long-form content, thereby increasing productivity.

The auto-generated commit summary feature in GitHub could revolutionize development workflows. 

A forecast by Forrester Research predicts that AI adoption in software development processes will grow by 40% by 2026. 

Thus, eliminating manual inputs could significantly streamline workflows, allowing developers to focus more on coding and less on administrative tasks.

Automation: HubSpot's integration with OpenAI

HubSpot's offers a new realm of interaction within CRM, whether it's processing leads or handling data. For instance, when a new lead submits a contact form, GPT-3 can instantaneously understand the context, generate an appropriate response, and send it to the lead, reducing the time taken by traditional manual responses.

Moreover, the AI can adapt previous user-generated templates for diverse scenarios, increasing efficiency while maintaining personalization.

And HubSpot isn't alone in this endeavor. We're witnessing an AI-infused CRM trend across the industry. 

For example, Salesforce's Einstein AI provides predictive lead scoring and intelligent insights, while Zoho's Zia uses AI to help businesses analyze customer sentiments in real time, enabling timely and effective responses.

AI-powered prioritization: streamlining customer support

HubSpot's AI now prioritizes and summarizes customer requests, interpreting sentiment and urgency from each interaction. Representatives no longer have to manually sift through a deluge of emails each day. Instead, the AI flags messages with detected urgency or negative sentiment, ensuring that pressing issues are swiftly addressed, enhancing the customer experience.

The future of ordering automation

With the recent advancements in language models, such as LLM and ChatGPT, the landscape of customer service and order automation is poised for transformation. Imagine walking into a McDonald's and interacting with an AI-powered kiosk that not only efficiently processes your order but also provides personalized recommendations, much like a human cashier might.

As illustrated in our simulation, these models can mimic the nuance and adaptability of human conversation, ensuring customers feel understood and catered to. As these technologies become more prevalent, we can anticipate a future where the boundary between human and automated customer service becomes increasingly blurred.

Сurrent stage of automation among the food industry leaders

With aquision of Apprente, a start-up specializing in voice-based, McDonald's aims to automate its drive-thru ordering process. This AI-powered system can understand multiple languages and accents, decipher complex orders, and operate seamlessly in noisy environments. However, is not devoid of initial teething problems. Customers have been quick to share humorous anecdotes on social media platforms like TikTok, recounting instances of misunderstood orders. 

Taco Bell recently showcased a concept restaurant featuring four drive-thru lanes, with orders delivered via a 'vertical lift' from the kitchen.

Starbucks utilizes its AI program, "Deep Brew", to create personalized customer experiences, manage inventory, and assist in-store placement strategy.

This AI solution also automates tasks like preventive maintenance on IoT-connected espresso machines. As a result, Starbucks has transformed into a data-driven company, leading to significant growth (for example, a 6% increase in same-store sales growth in the U.S.).

According to a McKinsey study, automating customer interactions could reduce the time taken to serve a customer by 20%, drastically improving service efficiency.

Further, envision a future where browsing through menus becomes redundant. With AI advancements, customers could potentially communicate their preferences via voice or text.

Researching complex information online

AutoGPT, as an evolution of OpenAI's GPT models, brings remarkable advancements in conducting complex research tasks and more. This AI tool can understand nuanced queries and delve into a diverse range of topics with impressive depth, generating relevant, insightful, and well-structured responses. It's capable of cross-referencing multiple sources of information within its training data and providing synthesized insights.

In the finance sector, a multinational investment bank, Global Finance Corp., integrated AutoGPT into their equity research division. The AI tool was able to quickly analyze and summarize large volumes of financial data and market reports, generating insights on investment opportunities and risks, which significantly enhanced the efficiency and depth of the bank's equity research.

AI's grasp on routine tasks

The general consensus is that tasks traditionally completed by individuals without specific qualifications can now be undertaken by AI systems. Companies are increasingly delegating tasks typically performed by junior-level employees to AI.

For instance, JPMorgan Chase & Co., a leading global financial services firm, employs an AI system named COIN (Contract Intelligence) to analyze complex legal contracts – a job that would normally take thousands of human hours. This allows junior employees, who previously spent hours reviewing these documents, to focus on tasks that require critical thinking and strategic input.

Natural superapp: a single app that gives users access to various types of services

Brain Technologies, the company mentioned earlier in our handbook, behind the newly-launched app "Natural", has secured over $50 million in funding, attracting a diverse pool of investors.

To use Natural, users either speak or type their queries or commands into the app. The app then processes these inputs and provides a solution, which could be linked to other apps the user might have.

For instance, if a user states "I'd like sushi tonight", the app might respond with options for ordering sushi from a variety of restaurants, recipes for making sushi at home, or places to eat sushi, providing a comprehensive solution. This works across various domains, including travel, food, and more.

Breakthrough in productivity management

Microsoft reinventing search with AI-powered copilot. Copilot is more than just a chatbot. It’s an AI assistant which gives different ways to command Office apps as well as a web browser. It will change forever how we create documents and interact with software. In the Office apps, Microsoft Copilot can generate the entire document for you as well as give precise recommendations to improve your text by introducing UI prompts that highlight your spelling mistakes. Copilot even can help you write emails in Outlook and so much more. 

Speaking of the web experience, you can chat with the copilot side-by-side as you browse.  It can help you turn your ideas into images in chat, refine your search results by asking questions, get summarized answers to search results to save your time, as well as generate content for you from scratch. 

Microsoft Word evolving from a basic word processor to a comprehensive writing assistant, offering style suggestions, text predictions, rewrite options, and a 'read aloud' function. 

And combined with Microsoft Copilot, Microsoft Word users can now manipulate and analyze data within their documents, leveraging AI capabilities to extract insights, generate reports, and automate data-related tasks.

Microsoft PowerPoint now automatically designs professional presentations, generates relevant slide content, and provides dynamic, personalized suggestions for slide layout and design based on the context. Additionally, the software offers real-time closed captioning and translation during presentations, accommodating diverse audiences.

Microsoft Excel's AI capabilities now include the Ideas feature, which automatically recommends charts and helps identify trends in data, along with advanced data analysis tools such as automated forecasting. Excel also provides rich data types, converting simple text into something much more meaningful, and the XLOOKUP function to find and retrieve data.

Combined with Microsoft Copilot, Excel's potential expands further. Users can command Copilot to prepare complex data visualizations, perform intricate calculations, or analyze data in new ways. Copilot's capabilities effectively turn Excel into a robust data science tool, accessible to users without advanced technical expertise.

Microsoft Outlook, coupled with Copilot, is revolutionizing email management. Copilot's AI assists with organizing the inbox, highlighting important emails, and even drafting responses based on previous communication styles. It can also extract actionable tasks from emails, such as scheduling meetings or setting reminders, optimizing productivity and reducing the chance of missing critical information. Furthermore, Copilot can provide brief summaries of long email threads, making it easier to catch up on ongoing conversations.

Microsoft 365 Copilot X enhances productivity in Microsoft Workspace by intelligently managing tasks across apps. It can delegate work in Teams, suggest OneDrive documents, and convert Teams meeting points into Planner tasks. It even provides automatic responses to messages, improving communication efficiency and fostering a collaborative work environment.

Jupiter Agent's integration with the Zapier Toolkit offers a transformative experience for programmers. Jupiter Agent, with its AI capabilities, assists in identifying code patterns, debugging, and even predicting developmental issues, while Zapier ensures a seamless workflow among the tools commonly used by developers. For instance, programmers can set up automated tasks like code backups, synchronizing project management tools with code repositories, or even auto-generate bug reports. 

Revolutionizing collaboration with OpenAI-powered Slack assistant

Slack isn't just a platform for team communication anymore; it's now a tool that assists with your daily tasks. The integration of OpenAI into Slack transforms the way we work and collaborate. It gives you an AI assistant that understands your queries, automates responses, and even schedules your meetings.

For instance, in Slack, OpenAI can help manage your tasks, provide quick information without leaving your workspace, and even suggest responses to messages, saving you typing time. This AI assistant doesn't just react; it understands and learns from your interactions, providing increasingly accurate and relevant responses.

Acceleration of music production: WavTool

WavTool is a unique AI-accelerated music production tool to create high-quality music and audio in a browser. It serves as a comprehensive solution for beginners, musicians, and producers alike, consolidating recording, composing, producing, mixing, mastering, and exporting functionalities into a single platform.

This Digital Audio Workstation is powered by a text-to-music AI assistant that has a solid grasp on music theory and audio production concepts and can clearly understand your ideas and try to implement them and help you with creating your music.

This tool has the ability to generate beats, offer chord suggestions, and generate melodies to kickstart your creative process. By leveraging AI technology, WavTool utilizes plain English input to explain concepts, provide suggestions, and guide you throughout the music-making process, ensuring a seamless and effortless experience.

The major shift in data visualization and generative AI

Generation, editing, and improving data

Large language models are now capable of creating all sorts of visualization: tables, charts, and other formats.

For instance, before AI-powered tools, you had to add custom widgets to a dashboard and each time you add something, there was a need to fill in a lot of fields. Now, you can simply feed the data to such a tool as WolframAlpha and choose the way you want it to visualize it.

When combined with a Code interpreter or the Notable plugin, ChatGPT demonstrates how effortlessly LLMs can handle data analysis. It can determine what needs visualizing, select the optimal visualization tool, and render the visualization effectively.

LLMs like GPT can write code across numerous languages and frameworks, vastly expanding their utility. 

Here is an example of the creation of visualized data from a text prompt. GPT 4 is able to create HTML, CSS and JS code to visualize required information including information from user input and information it has in it’s memory or can get from online sources:

The current stage, SVG and other frameworks

Technological advancements in data visualization are redefining the ways we represent and interpret data. Groundbreaking tools like Mermaid, Sketchviz, PlantUML, and Draw.io have emerged as frontrunners in this revolution, each introducing unique, user-friendly methods for depicting intricate data. Mermaid stands out with its markdown-like script language, simplifying the process of generating charts and making modifications almost effortless. Sketchviz, drawing upon the capabilities of Graphviz, transforms simple text files into detailed graphs, a testament to its emphasis on user-centric design.

In parallel with these tools, generative models like GPTs are pushing boundaries further. These advanced models can generate code that integrates seamlessly with visualization platforms, including the aforementioned Draw.io.

While current adaptations function via plugins, facilitating a "plug-and-play" experience, the future holds even more promise. We can foresee a world where visualization tools come equipped with built-in chatbots, powered by models like GPTs. Users will be able to simply convey their visualization needs, and these intelligent chatbots will craft the requisite graph, chart, or diagram instantaneously. This evolution is not merely about enhancing efficiency but about democratizing data visualization, ensuring everyone, regardless of their technical skill level, can harness the power of visual data representation.

Draw.io offers a comprehensive suite of diagramming features, supporting a multitude of export formats and integrations, which broadens its application scope.

Advancements in visual analytics

Interactive storytelling is more than just presenting data in a static chart. It involves creating a captivating narrative around datasets, with users actively engaging in the exploration process.

For the AI community, Chat GPT's new plugins, Code Interpreter or Noteable, are trailblazers in this sense, allowing third-party incorporation. This technology enables data uploading and analysis while also facilitating the output of the results in both narrative form and visual illustration. It mixes capabilities of LLMs and Python code interpreters with a reach toolset to analyze and visualize data. You can learn more about it here: 

Cieden recommends:

ChatGPT plugins | customer segmentation, clustering & targeting using...
How to do data analysis using Al with ChatGPT and the Notable plugin...

Let's walk through an example of such interactive data visualization in an application's user interface. Let's think of some app that helps users understand crime hotspots in a locale. A traditional approach would involve statically displaying crime data on a 2D map. 

Oakland crimespotting

However, with a code interpreter, this translates into an engaging exploration. 

First, users are guided about how such analysis of potential crime trends should be conducted, then the system may generate charts and textual interpretation over a past period, sifted from a variety of data points. As users delve deeper, they are greeted with detailed interactive visuals, like crime intensity mapped over the locale or timelines representing spike hours.

The story does not end there. The system could also provide crime resolution rates and types that dominate a particular area. It could go as far as clustering areas based on crime type and identifying patterns that could shape future prevention strategies. 

Dynamic widgets and reports

AI has begun to revolutionize dynamic widget and report creation in recent years, offering a shift from static, one-size-fits-all solutions to highly personalized, interactive, and real-time experiences. In 2023, we see this transformation gaining traction in several advanced products and applications.

An innovative product to consider is Databox, which uses AI to build highly customizable dashboards. It uses predictive analytics and machine learning algorithms to provide real-time insight, allowing businesses to make data-driven decisions instantly. 

Then there's Power BI's AI-powered features like Q&A, Quick Insights, and the Key Influencer visual, which enable users to interrogate their data naturally, identify trends and outliers quickly, and identify the main factors that drive outcomes, respectively. These AI functionalities not only improve efficiency but also provide in-depth insights that would otherwise require expert data analysis.

AI-driven visualization tool that seamlessly adjusts to users needs

Picture an AI-powered dashboard that seamlessly adjusts to your needs, integrating widgets with real-time data and selecting the optimal visualization method. Whether you favor tables or pie charts, the decision is in your hands.

Furthermore, such an AI assistant might detect trends and patterns and share insights that could potentially lead you to informed business decisions. If you were fine-tuning lead sources, this assistant could pinpoint situations with a better chance of closing successfully, identify leads of greater potential value, and highlight opportunities for upselling or cross-selling. Beyond that, it might suggest actionable strategies to potentially enhance sales and revenue.

Final thoughts

Design's role in modern user experience

In the mid-1960s, the advent of time-sharing marked a significant shift from the batch processing paradigm, introducing command-based interaction. This paradigm transition witnessed users and computers in a collaborative dance of commands across varied platforms, from command lines to the more advanced graphical user interfaces (GUIs). With their visually rich display of system statuses, GUIs provided users with an enhanced two-way interaction, underscoring their usability across various applications and solidifying their dominance in the tech landscape for over six decades.n.

Fast forward to today, and the rapid evolution of technology presents a fresh twist in the narrative of user interfaces, championed by the AI revolution.

This new era, defined by "intent-based outcome specification," sees users and AI in a more dynamic relationship. Instead of meticulously instructing computers, users now state their desired results, and generative AI tools like ChatGPT take the reins from there. However, this isn't a silver bullet solution. Many users grapple with articulating complex requirements, and these tools are not devoid of limitations. 

Consequently, the future seems poised for a hybrid interface—a blend of command-based interaction's tactility with the advanced intent-recognition of AI. Reflecting on this evolution, we notice an emergent trend among successful businesses: the increasing emphasis on UX/UI design. According to a PWC study, a staggering 90% of top-performing companies recognize design as an indispensable component of their strategic foundation.

If you're eager to lead the charge in this exciting frontier and want expert guidance, reach out to us. We're more than ready to collaborate and bring your vision to life!