Anticipating AI's next move ? article ① ?

If you are a regular reader of TTS articles, you know I usually aim for a reading time of well under 10 minutes. I get it - the attention spans are short these days

Aye, I be soundin' old.

But the topic I'm diving into today, tomorrow, and the day after is a bit more complex then all that came before, and so it will take a bit longer to cover. So I apologize in advance, but knowing that TTS readers are both highly intelligent and hungry for deep cut insights, I decided it’s worth going the extra mile.

To make it more accessible to everyone, not just the most dedicated, die hard readers, I've decided to cut it up into three pieces.

Each day will feature the next piece of the article. And if you don’t want to wait for tomorrow, you can read the entire article on my blog TechTonicShifts (please note, the site is still under construction).

This article introduces a framework that I created some time ago to help anticipate future developments in the AI/ML space. And like any framework, it gives a 50.000 feet overview, focusing on the big picture rather than every single development.

This framework is a visual aid or infographic to help predict the next steps in AI's evolution. It is not exhaustive, but it is meant to guide your understanding of where AI is headed.

Because I’ve been working with this framework for over a year, many of the articles I’ve written have been inspired by it. The great thing now is that I can finally reap the benefits of all that work. Articles that are relevant to specific subchapters are mentioned, and you can link back to the original content.

So, without further ad…. Ahhh.. nah.. I hate this phrase.

Just get on with it!

Before we start!

If you like this topic and you want to support me:=

1. Comment on the article; that will really help spread the word ??
2. Connect with me on Linkedin ??
3. Subscribe to TechTonic Shifts to get your daily dose of tech ??

The stuff from science fiction

When I was first introduced to real Machine Learning, it was 2007. Before that it was just stuff of Isaac Asimov's science fiction dreams or Arthur C. Clarke's dystopian nightmares.

I began to imagine a future where intelligent machines could work alongside humans to revolutionize our lives and reshape society. That future, it turns out, is no longer a distant fantasy.

From the moment I first encountered technologies like Simple Recurrent Networks, Word2Vec, and NLP for simple tasks like word prediction, I knew we were on the verge of something new.

But I've noticed from the conversations that I am having, that for a lot of people, the idea of AI still feels more like a plot deviced from a Terminator movie than a practical tool for driving growth and innovation. To them the concept still seems daunting, the technology too complex, and the implications too uncertain.

That's precisely why there has been a slowdown in the AI hype, with AI stocks plummeting, and business leaders shifting their focus from building GenAI gadgets to prioritizing tangible outcomes.

If you happen to find yourself in this camp, you should not worry – you're not alone.

Because embracing AI doesn't require a PhD in computer science or a crystal ball to predict the future. What it does require is a clear understanding of how AI can be integrated into your life or your business, a roadmap for navigating the challenges and opportunities that come with it, and of course a vision for how it can transform your environment for the better.

And hat's where this guiding framework comes in.

About the framework

I have broken down the evolution of AI in business into a simple, straightforward framework so that any human; a regular John Doe, a technology architect or a CEO could easily understand and apply.

I intended this framework to help guide you from the basics of AI-powered content generation to the cutting edge of AI-First business redesign and human-AI augmentation, all the way to the "singularity".

And along that way, I'll demystify the jargon, cut through the hype, and provide some examples of how others are using AI.

The journey starts in the Generation Space. Here, AI handles content creation, data analysis, or sentiment analysis. Next is the Productivity Space. AI becomes a key driver of efficiency here. Autonomous tools manage customer inquiries 24/7. Predictive analytics forecast demand, and you carry your personal shopper with you wherever you go.

Here, AI isn't just helpful; it's transformative.

Because it reshapes how you or your business operates.

Click here to warp to the Productivity Space ....

To me, where AI is disruptive and a real game-changer is the third space, where AI-Native Business (re)Design takes place. AI takes center stage in this domain. AI-first businesses are driven by algorithms and supported by robots. They redefine what a company can be.

And lastly I'll be focusing on the singularity, the moment when AI turns into something way more intelligent than we as humans are. And that is called Augmented General Intelligence, and this will happen because of developments in AI models, and also Quantum Computing.

AI will be integrated into robotics (embodied AI), and humans will become augmented by it. And this will lead to the most profound revolution in the human history. The singularity is not a futuristic scenario though. Key developments have been long underway, and some expect us to reach AGI around 2029.

Some people are worried about AI replacing jobs. That is rightly so, but the future is also about collaboration. Human-AI partnerships amplify each others' strengths. They combine AI's analytical power with human creativity and judgment.

It does not matter if you are a seasoned tech veteran or a curious newbie dipping your toes into murky AI waters for the first time. I have written it in such a way that this guide will provide you some insights you need, to survive the coming AI-driven disruption.

Now, with that said, let's begin with the Generation Space.

The Generation Space: the first baby steps into the world of (generative) AI

Welcome to the Generation Space, the first stop into the world of AI's evolution.

This is where it all began for most of us, with OpenAIs heroic achievement, launching ChatGPT3. The Generation Space is the place where artificial intelligence, as we have come to know it, took its initial, tentative steps into our world.

If the Productivity Space is where AI learns to walk and the Business (Re-)Design Space is where it starts to run, and the Singularity Space, where AI starts to outrun us, then the Generation Space is AI's infancy – a time of exploration, experimentation, and rapid learning.

For many people, and companies alike, the Generation Space is the most accessible and least intimidating entry point into the world of AI. It is a space where intelligent machines are tasked with generating ideas, spotting trends, and handling the kind of repetitive, data-driven work that we as humans find tedious and time-consuming.

Even though I had known for some time that AI would make huge waves, I was as surprised as anyone when ChatGPT came along. Sure, like many others, I had been toying around with previous models from OpenAI, like GPT-2 (2019) and GPT-1 (2018) – playing with it in Python, along with their rivals like TensorFlow (Google Brain, 2015), and PyTorch (Facebook, 2016). But none of us anticipated the revolutionary leap that GPT-3 would make—surpassing even the expectations of its creators.

So what does AI look like in the Generation Space?

Let's get to work:

Content creation

One of the most common applications of AI in the Generation Space is content creation.

This is where tools like OpenAI's GPT-4o for omni (text, images, coding, analysis, visualizations), Claude Sonnet (text, analysis, visualizations), LLAMA (text, images, coding), Midjourney (Images), Suno (Music), and others come into play.

All of them use the power of natural language processing and neural networks to generate human-like text, images or music with minimal input. Whether you need a blog post, a product description, a social media update, or an email campaign, the AI will take your basic ideas and transform them into a polished, and engaging copy that reads like it was written by a skilled human wordsmith.

Well, up to a degree, that is….

If you've read my article Don't want to be caught using ChatGPT? Than stop delving ! , you will have learned how to smell a GPT generated rat from a mile away.

The beauty of AI-powered content creation is that it saves your team enormous amounts of time while still delivering reasonably high-quality results. Instead of spending hours agonizing over every word and phrase like we used to, we can now focus on the big picture—strategy, tone, and messaging—while AI handles the tedious work.

Of course, AI isn't just about generating text – it can also create images, videos, and even music with startling proficiency. Tools like Midjourney and Suno are pushing the boundaries of what's possible with AI-generated visuals and music. They are producing stunning works of art that blur the line between human creativity and machine learning.

Duhh.. I just realized the sixth finger when typing this praising paragraph.

But do not dismay, the genAI company Ideogram promises us that they excel in rendering human features and text within images which, as we all know, has been challenging for AI models like Dall-E, and Midjourney/Leonardo.

Data analysis - insights at the speed of light

Another key application of AI in the Generation Space is data analysis.

Businesses are drowning in information – from customer data to market trends to social media chatter. And when you, as a sales leader, want to have the monthly sales figures, offset against a certain variable which is not in the standard reports, you will have to wait a couple of days before you can take action.

For those people, AI is a true lifeline.

It is helping companies make sense of the chaos because it can extract valuable insights that can drive better decision-making. Because if you have a tool that can sift through millions of data points in seconds, you are able to identify patterns and correlations that would take a human analyst weeks or months to uncover.

Now that is the power of AI-driven data analysis. With Machine learning algorithms, can get a deeper understanding of our customers, our markets, and our own operations – and all in real-time.

Take For instance Anthropic, a competitor of OpenAI, with their latest model Claude Sonnet 3.5. You can use it for writing, but where it really stands out, is its insane capability of generating interactive dashboards. These interactive dashboards are transforming data from earnings reports, research papers, and whatever you can think off, into user-friendly visualizations

The video below shows how Claude AI can help you visualize your data and turn it into beautiful reports.

I have built predictive analytics tools in the past that can help companies forecast demand for products or services with a bloody good accuracy. One of these predictive tools for instance was for an in-company-restaurant-chain visitor-forecasting-system (…that could be done shorter !) for a company called Recruit.

By predicting the amount of visitors given historic data, and taking into account variables like the weather, the menu, days of the week, etc., they were able to forecast within a certain limit, how many people would come visit. This would allow them to optimize their staff, and waste less food.

With predictive analytics I even went as far as trying to analyze how energetic I would be feeling the next day. By monitoring glucose levels, current energy score, sleeping patterns, air pressure, caloric intake, heartrate etc., I was able to notice that there is a positive correlation between some of these variables and my personal welbeing the next day (read the article on digital phenotyping )

Below is an image of a multivariate-analysis of different variables that influence my energy level the day after.

But also think of social listening tools. They can monitor online conversations about a brand or industry, and are providing you with early warning signs if a potential crisis is erupting, or if an opportunities arisis.

And what about customer segmentation tools. They can identify distinct groups of customers based on their behavior, preferences, and needs, allowing businesses to tailor their marketing and sales efforts for maximum impact (read the articles on hyper-personalization, here and here ).

The key advantage of AI-driven data analysis is speed.

And that is because human analysts might spend days or weeks poring over spreadsheets and reports, while AI can crunch the numbers in seconds. And that is the key difference between giving businesses a critical edge in fast-moving markets where timely insights can mean the difference between success and failure.

Trend spotting - seeing the future through AI eyes

Another powerful application of AI in the Generation Space is trend spotting.

I hate the sentence below (because it is a tell tale sign that you are dealing with an AI generated text), but in this case it is the correct phrase to use….

In a world that is changing faster than ever before – where new technologies, consumer preferences, and market forces are reshaping the world as we know it – just trying to stay ahead of the curve means survival.

But how do you anticipate the future when it's always in flux?

The answer is obvious, given the context of this article:

Artificial Intelligence

Like with the case of Claude, the AI is capable of analyzing lots of data from a wide array of sources. From unstructured data like social media, news outlets, industry reports and more, to the structured stuff that lives inside databases, datamarts, and lakehouses.

Algorithms can identify emerging trends and patterns long before they become apparent to us humans. This is achieved through the use of unsupervised pattern recognition algorithms, such as clustering algorithms and anomaly detection methods like K-means, or Autoencoders, in combination with things like graph databases (read the articles: Knowledge graphs for Machine Learning are so cool ! and Why vector- and graph databases are so cool for AI )

These algorithms analyze vast amounts of data without needing explicit labels, and they uncover hidden structures and relationships that can signal the onset of new trends or the emergence of unexpected patterns. And this in turn gives businesses a beautiful head start in preparing for a change in their market, so they can developing new products or services, and stay a little bit ahead of the competition.

For example, a fashion retailer like Zara uses AI to analyze social media posts and online search data to identify emerging style trends months before they hit the mainstream. Zara adjusts its product lines, marketing campaigns, and inventory levels to benefit first from the coming wave of consumer demand.

Similarly, a technology company like Google uses AI to monitor patent filings, academic research papers, and startup activity to identify promising new areas for investment and acquisition. This way they are staying ahead of the curve, and Google can position itself as a leader in these emerging markets and technologies, rather than playing catch-up to more nimble competitors.

They have done this in fields such as AI, quantum computing, and autonomous vehicles.

Cool innit?

Of course, trend spotting is more than just a defensive tactic – it can also be a powerful tool for innovation.

When Netflix used AI to identify unmet needs and untapped opportunities in the market, they developed entirely new services like personalized content, and original programming. This led to the creation of their highly successful original series and movies.

This strategy allowed Netflix to transition from a DVD rental business (bet you didn't know that! if you are not from the US) to a global streaming giant. And with that, they even created a new business model that disrupted the entire industry.

The Generation Space is where it all begins – the first steps on the path to AI integration (the productivity space) and transformation (the third, and disruptive space).

For businesses that are new to the world of AI, this is often the most accessible and least intimidating place to start. And you can see that based on the number of chatbots that are popping up everywhere. Whether you like them or not (read the article: We don't want to talk to your chatbot )

When you start using AI for content creation, data analysis, and trend spotting, you are dipping your toes into the water and start to see the potential for intelligent machines to drive real business results.

But of course, the Generation Space is just the beginning.

Because when businesses become more comfortable with AI and start to see its benefits firsthand, or when people see the huge investments and need to see results (see the article: The generative AI bubble ), they naturally want to explore more advanced applications and integrations that pay off.

And that's where the Productivity Space comes in – a place where AI starts to take on more complex tasks and becomes a true driver of efficiency and innovation within an organization, or with you as a person.

More predictive analytics - using historical data to predict future events

As mentioned before, predictive analytics is a crucial element of the Generation Space.

With these Machine Learning tools, companies begin to truly see potential of AI by forecasting future events.

Traditional analytics focuses on understanding past events, but predictive analytics uses historical data and statistical algorithms plus machine learning to forecast what will happen next. And by doing this, these businesses go from reactive decision making to proactive.

Take Netflix. They use Amazon Forecast to analyze user viewing history and predict what movies or TV shows you would likely enjoy next. And by comparing your preferences with the collected data from millions of other viewers, they can highly personalize its recommendations to keep you hooked.

Predictive Analytics has been part of the Netflix algorithm from the start and that made them who they are now. Of course Amazon was the first big company to experiment with this technology and they reaped the fruits of it years ago.

Also Walmart is another beautiful case of predictive analytics in action. The retail chain uses SAP Predictive Analytics to forecast product demand across its network of stores. And for that to happen, they analyze a blend of historical sales data, local events and even weather conditions. With this, Walmart can predict which products will be in high demand in specific locations.

And with this foresight Walmart gets to optimize its inventory to make sure that popular items are always in stock. It boost sales, and it reduces waiste!. This capability is especially critical during peak shopping seasons, where demand fluctuations can significantly impact revenue.

UPS also uses the power of predictive analytics through its ORION (On-Road Integrated Optimization and Navigation) system. ORION analyzes data from over 250 million package deliveries to predict the most efficient delivery routes. They factor in variables like traffic patterns, weather conditions, and delivery times. And by collecting this data, UPS anticipates potential delays and adjust routes in real-time.

UPS uses a combination of two algorithms to accomplish this - a "traditional" combinatorial optimization algorithm (heuristic, and linear programming algorithms), that solves the so-called 'Salesman' problem. The Traveling Salesman Problem has been around for a while. It is a classis optimization problem where to goal is to find the shortest route possible that allows the salesman to visit customers exactly once and in an optimal way.

They use this classic algorithm in combination with Predictive Analytics that looks at the road, weather, etc. The combination is so powerful that they save between 400-500 USD annually with ORION.

In my work with the Public Employment Service in 2012, I applied predictive analytics techniques, in particular logistic regression and decision trees, to implement a dynamic profiling system for job seekers. This approach allowed the PES to continuously assess and update each job seeker's risk of long-term unemployment based on a range of factors such as demographic data, work history, social heath determinants and local labor market conditions.

And by using these techniques, I could generate a dynamic profile for each individual that reflected their likelihood of securing employment within a given timeframe. This real-time profiling enabled us to segment job seekers more effectively than before. This lead to people, with a higher risk of unemployment, receiving the targeted interventions they needed. The dynamic nature of the profiling meant that as job seekers' circumstances changed, so did their profiles. And this allows for a more personalized and responsive service delivery.

In conjunction with dynamic profiling, I used an optimization algorithms; the Hungarian algorithm and linear programming for Outcome Based Referral (like UPS). These algorithms were crucial in matching job seekers with the most suitable employment services and opportunities, based on their dynamic profiles.

The goal of this second step called Outcome Based Referral was to optimize outcomes so that job seekers were referred to the best PES services (called Active Labor Market Policies) that would most likely result in successful job placements, against the lowest cost for the budget issuer.

And by incorporating factors such as service availability, job market trends, and individual job seeker profiles, the optimization algorithms helped maximize the effectiveness of our referral process.

This outcome-based referral system improved job placement rates significantly and also allowed for an optimal allocation of PES resources (services, budgets, etc). Overall these three algorithms reduced the time job seekers spent unemployed and it improved the overall performance of the Public Employment Service.

In the healthcare sector, you have the famous Mount Sinai Hospital in New York. They use IBM Watson Health for predictive analytics to improve patient outcomes. It analyzes patient records and genetic data, and even social determinants of health (like housing, income, etc.).

Watson can predict which patients are at higher risk for complications or readmission. This allows doctors to intervene earlier and tailor the treatment plans to the individual. The hospital’s use of predictive analytics has not only improved patient care but also reduced healthcare costs.

Wrapping up the Generation Space

Now that the exploration of the Generation Space is done, it is clear that this foundational phase in AI's evolution is where many businesses will begin their AI journey.

Here, AI is primarily about generating content, analyzing data, visualizations, and predicting future events. It is a space where AI flexes its muscles and helps people and businesses with automation and with simple task execution that were once time-consuming and labor-intensive. From predictive analytics in the healthcare sector to the dynamic profiling of job seekers in public employment services, the Generation Space is setting the stage for more advanced AI applications.

But this is just the beginning. In the second episode, that will feature tomorrow, I will discuss the second space —the Productivity Space— where AI moves from assisting with tasks to fundamentally driving efficiency and innovation across entire business processes, and helping people with the automation of mundane activities and chores. Here, AI doesn’t just help; it starts to transform the way we live andwork.

If you want to jump to the second article: the Productivity Space - just click here.

In the third space, the day after, I will touch on the Singularity Space. That is where AI's potential reaches its zenith, and we begin to contemplate a future where AI might even surpass human intelligence. This is the cutting edge of AI, where the possibilities are both exhilarating and, for some, a bit daunting.

If you have come this far, you are a die hard geek as I am. And I thank you from the bottom of my heart for sticking around with me for so long.

And I have something for you:

Solve this little math puzzle, and get a copy of my latest book

Exploring consciousness: a guide for AI students

What is: (8 x 3) - 6 / 2?

Post your answer in the comments below and I'll contact you!

Stay tuned as I explore this exciting space in tomorrow’s new episode.

Signing off - Marco

Well, that's a wrap for today. Tomorrow, I'll have a fresh episode of TechTonic Shifts for you. If you enjoy my writing and want to support my work, feel free to buy me a coffee ??

Think a friend would enjoy this too? Share the newsletter and let them join the conversation. LinkedIn appreciates your likes by making my articles available to more readers.

Links from the article:

• How I forecasted my energy levels through Machine Learning - Digital Phenotyping.
• The age of hyper-personalization
• Combining machine learning with hyper-personalization
• Knowledge graphs for Machine Learning are so cool !
• Why vector- and graph databases are so cool for AI
• We don't want to talk to your chatbot
• The generative AI bubble
• Don't want to be caught using ChatGPT? Than stop delving !

Top-rated articles:

• Fount helped save 13.4M in nine months with AI based work optimization.
• Prepare for the "AI-first" businesses.
• AI based Intelligent Automation: [RPA + BPM] x AI = ROI
• A model for strategy development in the AI age
• We don't want to talk to your chatbot
• Copyright infringements by all major AI players
• Lobotomized AI*
• The AGI hype = Bored Apes
• 150 Digital Transformation templates new ones added
• Chip happens ! This article explains the alphabet soup of chip names (CPU, GPU, LPU, TPU)
• Now why do AI/ML project FAIL!
• The age of hyper-personalization