Skip the hype (and yes there’s a lot of it out there right now.) I’ve spent the better part of this year speaking to people gathering their thoughts on the “birth” or “coming to light” of Artificial Intelligence (AI), and pieces of those conversations are reflected in this article. My general conclusion is that there’s nothing to fear right now. You shouldn’t be expecting a Terminator army coming for you, yet. So, let’s dump the hype and get right into it.

What is artificial intelligence? Artificial means “made or produced by human beings rather than occurring naturally, especially as a copy of something natural.” Ref: Google’s English Dictionary provided by Oxford Languages. So, something that does not occur naturally in nature…and the definition of intelligence per the same reference is “the ability to acquire and apply knowledge and skills.” Now we have a challenge. Does AI as you perceive it have the ability to acquire and apply knowledge and skills? Not in my world; it’s still a bunch of algorithms and lines of code set to do certain things, however with a bit more leeway than your Basic, Fortran, Cobol, C++, Perl, Python or any other code will give you. Nothing is solid or remains in a constant state with evolving technologies, everything moves, morphs, and rarely stays the same for long. Technology is malleable, and will bend if smart people push it.

The realm of AI is an ever-evolving landscape that has captured the imagination of people worldwide; its potential seems limitless: from revolutionizing transportation with self-driving cars to enhancing daily life through voice-activated assistants, and efficient data mining. I joined a group of volunteers to help put on the DecodingTech.Zone event that highlighted AI in Brampton, Ontario October 2023 (full disclosure: Betweenplays Media was the official media sponsor for the event), and that also lent itself to be an eye-opening experience, and further contributed to this article. Have a look if you’re intrigued.

Look carefully. Listen widely. The event can be seen on live stream here (I’m in there somewhere—the only person with a red shirt):

The “live from the floor” interviews can be seen here, Live from the floor!

But I digress. Getting back to the core of the discussions I’ve had…

What is available right now isn’t new, it’s been around for a long time, and organizations have been playing with AI for a long time. What you see today is what is generally targeted at end users and has a commercial model built for the public. I think there’s a crucial aspect of AI that often escapes the spotlight: the ever-elusive commercial model for everyday enterprises. Meaning that right now it’s a toy for you and I to play with, invite to our video calls to make things a little more efficient, and draw a bunch of end of the world “last selfie every” pictures. Enjoy that!

In my opinion, what we need is a viable revenue model that transcends beyond data lakes and mining our own information better than we could do on our own, and that’s not too far away, but it’s not quite here today. Yes OpenAI just released a bunch of “tame-able” babyGPTs that you can customize to do your bidding, but it’s really just a text based Dall-E/Midjourney queries that I don’t suppose is going to do much for now. With that being said, we need to fight the internal fears we have and what the regulators want us to believe is possible with AI and temper it with what is possible right now. Strange days ahead for certain, but no one is going to be showing up asking if you’re Sarah Conner anytime soon. Oddly though, to understand how to commercialize AI, we do need a time machine.

AI’s Historical Odyssey

The history of AI extends far beyond recent developments; it’s a tale that spans several decades. Conversations I’ve had with people in the know have revealed valuable insights. Nokia, Bell Canada, Verizon, Comcast and other telecos and media companies have been experimenting and building AI for over 30 years. Even the “suggest” function in Google’s search and “people who bought this also bought”, and the “recommended for you” are all rudimentary elements of AI. We just thought it was convenient, but it’s always been there. We just didn’t notice.

This lengthy history posits that not only is AI not a new concept but it’s been around for a very long time and it has been continually evolving field with deep-rooted origins. Remember IBM’s Deep Blue playing chess and Google’s (purchased) Deep Mind playing Go? This isn’t new so stop being afraid of what we’ve already accepted. There was an effervescence of artificial intelligence long before 2023, we as a people have been on this quest for a very long time; it’s almost like the invention of fire; it just doesn’t burn…or does it?

The Chief Revenue Officer’s (CRO) assertion is a pivotal one: if there were a viable commercial model for AI, it would have been harnessed long ago. The path to practical AI application in the commercial world is far from straightforward. Generative AI, a subset of AI, has existed for quite some time. Nonetheless, its recent emergence into the mainstream mirrors the impact of the Internet when it first became accessible to the public. It ignited a frenzy of innovation, resulting in the proliferation of websites. Yet, the question persists: What practical commercial applications does AI offer today? Art? Data analytics? I think the things we need it to do are still slightly out of grasp, but not so much so as this sentence purports. Now we find ourselves in a world where people are experimenting and trying to push AI forward. Take for example Dictador, a Polish Rum distillery that’s now elected to have a robot-ish AI as their CEO. The interviews with it are rough and lacking. I won’t bother putting the link in the article but you can find them. The interesting thing is that the article I did read says:

“Mika will largely focus on choosing artists to design custom bottles.”

That’s not much of a CEO’s job, and a little more like a marketing concern. 

OpenAI’s Arrival on the Scene

OpenAI, as a company, does not introduce a groundbreaking concept. Instead, it’s one of many organizations navigating the intricate maze of AI commercialization. OpenAI primarily revolves around a business model that charges retail consumers or end-users for their services. However, the practical business applications of AI remain limited, except in the realm of massive data mining. Even then, the process of training AI models is painstakingly time-consuming. I pick on OpenAI because it’s cheap and easy for me to do so, but like Apple back in the day they were first to market with he iPad, so that does give them an advantage. I play with ChatGPT every day. Albert (Betweenplays CEO) and I even interviewed Chat earlier this year. You can see that interview here:

Consider Deep Blue from IBM, which has been nurturing Watson, its AI counterpart, for over four decades. Deep Blue’s initial claim to fame dates back to the 1990s when it exhibited prowess in chess. 

Ref: https://www.britannica.com/topic/Deep-Blue

AI is not a novel innovation, but its allure and potential for transformation are undeniable. So where do we start, because believe me we haven’t even gotten started yet. We are but neophytes. Don’t get me started on Google’s Alpha Go Engine. If you haven’t seen the film, you should. Watch as AI brings a Go master to tears, it’s an incredible thing to watch:

There’s nothing like this, and then suddenly there’s a genesis where we all think this is new? No way. We’ve been working on this for a long time. SkyNet isn’t here yet, but it would have been if there was a way to make money off it, and we just haven’t figured that out.

“So You Basic”: LLMs v. AI

A large language model (LLM) and artificial intelligence (AI) are related concepts, but they have distinct meanings, and purport to do completely different things. They’re really not that dissimilar; I believe however, that they do function on different planes.

Large Language Model (LLM):

A large language model is a specific type of system that is designed to understand and generate human language; it is trained on large datasets of text and uses machine learning techniques to learn the patterns and structures of language. LLMs are typically used for natural language processing tasks, such as text generation, language translation, sentiment analysis, chatbots, and more. They can understand and generate text in various languages and are widely used in applications like virtual assistants, language translation services, and content generation.

Artificial Intelligence (AI):

Artificial intelligence is a field that covers the development of intelligent machines or systems capable of performing tasks that typically require human intelligence (but not all tasks). AI covers a wide range of techniques, including machine learning, deep learning, natural language processing, computer vision, robotics, expert systems, and more—let’s not forget marketing and productization, because without that you just have a mountain of code. AI systems can be designed for various applications, such as data analysis, decision-making, pattern recognition, problem-solving, and automation. AI extends beyond just language processing and includes a wide spectrum of tasks and applications.

Sadly, without a use case that can be commercialized you’ve got nothing. Retail AI needs to transform into Enterprise AI, or what I like to call Effective AI. So what are the use cases? They’re out there. I believe.

The Times To Come and The Ever Elusive Enterprise Use Case

AI’s true potential shines when it executes enterprise use cases that extend beyond mere data analysis. The power of AI becomes evident when it eliminates mundane tasks from employees’ daily responsibilities, freeing them to focus on higher-level functions. This shift empowers individuals to unleash their creativity and problem-solving abilities, which are often constrained by repetitive tasks like copy-pasting data between applications due to the lack of interoperability of applications.

Presently, AI in the consumer sector excels at creating slide decks, web sites, and making art or deepfakes. While numerous compelling commercial use cases exist, very few companies, if any, have succeeded in executing and productizing them on a scale that would make them commercially viable. The challenge lies in identifying the right enterprise use cases and effectively translating them into tangible products, and then the elephant in the room which is power, bandwidth, and demand. The last one is a killer, because without a revenue model it’s not sustainable, and you will lose money. I can almost bet that most companies running AI systems are paying more per kilowatt hour than they are making. 

The Swift Pace of AI Development

The excitement surrounding AI today, despite the old rot that it comes from, can be attributed to the rapid pace of development and innovation. Many companies are now actively promoting consumer-grade AI tools, but this phenomenon is not a spontaneous occurrence. These organizations have been quietly developing generative AI products (and others that we haven’t seen yet) for the consumer market for quite some time. Think of massive drone colonies moving together like a flock of birds. Yes, that’s already here. What do you think is the optimal way of controlling it. For now just humans.

Forget about A Flock of Seagulls, check out A Swarm of Drones:

OpenAI happened to be the first to make a significant impact, which led to widespread excitement and market / consumer interest (effectively mainstreaming the entire subject—which I support)—the Elon Musk factor helps too. This scenario is reminiscent of Apple’s introduction of the iPad, which was available a full year before any competitor released a similar product. Apple managed to catch its competitors off guard.

It’s essential to note that the lag in hardware production cycle compared to software development contributes to these occurrences. Software, when managed efficiently, yields rapid results, resulting in a software explosion, and for AI in the retail space, there is no time like the present. So where are all the big business ideas?

Unraveling the Complex World of AI Commercialization

AI’s journey has spanned decades, and its potential is being realized at a snail’s pace. The practical applications of AI in the commercial world are anything but straightforward, as its practicality remains limited in various sectors.

This is not just about having access to the technology; it’s about finding the right use cases, understanding the business and revenue models that will support them, and ensuring seamless integration into existing systems. Transition people. We are in transition! There are no such things as smart cities. Much like sex there is safer sex, not safe sex. So, we can build smarter cities and smarter devices, but there is no such thing as a smart city right now. They simply just aren’t smart. Yet, the quest for the perfect commercial model continues as AI reshapes industries, influences economies, and transforms the way we live and work, but no one is losing their job anytime soon; and to be honest the jobs that might be eliminated aren’t the ones that anyone really wants to do. We are all inherently lazy and want to do the least to enjoy ourselves the most. That’s my opinion and only the opinion of some of the people my year has led me to.

AI’s Role in the Evolution of AI Commercialization

By offering consumer-grade AI tools, the forerunners have opened new avenues for individuals and small-scale applications. They have been at the forefront of making AI accessible to a broader audience, and even putting it into the news – marketing and product not talking to sales. The age-old story.

While we look at better models to commercialize AI, it’s evident that current business models center around retail consumers and end-users. This strategic focus has allowed AI companies to lead the charge in introducing it to the masses. However, the business landscape for AI is not limited to consumer-grade applications. There are others out there focusing on training AI models for closed data loops “data lakes” and those models work. Once trained you can query across multiple companies owned by a single organization and ask, “What is our year over year revenue in this sector is across all the businesses we own?” You could do that with people but not in 1 minute, so there is a use, and it is viable; but is there more than just playing Hot Wheels with this thing?

The Business Motorcar for Enterprise AI

In contrast to the retail-oriented approach, Effective AI (So we’re clear I’m talking about what comes after Generative AI—it’s effectively chatter) focuses on harnessing the power of data to improve business operations and drive growth. Effective AI needs to seek to address specific challenges faced by companies, cities, and deserts by leveraging data analytics, automation, and intelligent decision-making processes.

Effective AI should aim to empower organizations by enhancing efficiency, reducing operational costs, and delivering data-driven insights; its applications range from automating routine tasks to optimizing supply chain operations, from improving customer experiences to enhancing cybersecurity.

AI for networking, and networking for AI. Such a conundrum. Infrastructure and how it’s managed is at the core of everything. Without network there is nothing, and no one can communicate with each other, including AI workings. They don’t work.

Successful Effective AI implementations will need to demonstrate tangible benefits, lead to cost savings, improved productivity, enhanced customer satisfaction, and more informed decision-making.

Realizing the Full Potential of Effective AI for Enterprise

The journey to harness the full potential of Effective AI involves several key steps and considerations. To deploy AI in an enterprise setting, companies must:

Identify Relevant Use Cases: Understanding the specific challenges and opportunities within the organization is the first crucial step. Identifying use cases where AI can make a meaningful impact is essential.

Data Integration and Quality: Ensuring that data from various sources is accessible and of high quality is a prerequisite for effective AI implementation. Clean, well-structured data is the foundation of AI-powered insights.

AI Model Development: Developing AI models tailored to the identified use cases is a critical phase. This involves training AI algorithms on relevant data and continuously refining them for optimal performance.

Integration with Existing Systems: Seamless integration of AI solutions with existing software and hardware systems is essential. Compatibility and interoperability are key considerations.

Change Management (transition!): Introducing AI-driven processes may require changes in organizational workflows and culture. Employee training and change management strategies are crucial.

Monitoring and Optimization: Continuously monitoring AI systems and fine-tuning them based on performance metrics is necessary. This ensures that AI solutions remain effective and relevant.

Data Security and Compliance: Safeguarding sensitive data and ensuring compliance with data privacy regulations is a paramount concern in enterprise AI.

Real-World Applications of Effective AI

Effective AI is not confined to theoretical possibilities; it’s a reality with tangible applications across various industries. Let’s explore some real-world examples of how Effective AI can transform businesses:

1. Customer Relationship Management (CRM): AI-powered CRM systems analyze customer data to provide personalized recommendations, automate follow-ups, and improve customer interactions. Salesforce’s Einstein AI is a prominent example.

2. Supply Chain Optimization: AI helps companies optimize their supply chains by forecasting demand, managing inventory, and enhancing logistics. Walmart utilizes AI to optimize its supply chain and reduce operational costs.

3. Fraud Detection: Financial institutions employ AI to detect fraudulent activities by analyzing transaction data and identifying anomalies. Mastercard’s Decision Intelligence is an AI-driven fraud detection solution.

4. Healthcare Diagnostics: AI is revolutionizing medical diagnostics by analyzing medical images, such as X-rays and MRIs. IBM’s Watson for Healthcare is an AI platform used for diagnosis support.

5. Chatbots and Virtual Assistants: Many businesses use AI-driven chatbots and virtual assistants to provide customer support, answer queries, and improve user experiences. 

6. Predictive Maintenance: In manufacturing and industrial sectors, AI predicts equipment failures and maintenance needs, reducing downtime and operational costs. General Electric’s Predix is an industrial IoT platform that incorporates predictive maintenance capabilities.

7. Natural Language Processing (NLP): AI-powered NLP is used for sentiment analysis, content summarization, and language translation. Google’s BERT (Bidirectional Encoder Representations from Transformers) is an example of a powerful NLP model.

8. Marketing Optimization: AI analyzes vast datasets to personalize marketing campaigns, target specific audiences, and improve conversion rates. 

Don’t even get me started on how we build memory and keep things under control. I have a model for that but that is truly a different article, and approach, and perhaps something I’ll open source once it’s thought through completely. More on that to come. If you want a hint think of blockchain and prime numbers. 

How to build Effective AI

So maybe it’s not about generative AI. I think it’s about creating Effective AI (you heard that here first people). Cradle to cradle is a biological approach to thinking / AI. This is the though process of use everything and then re-use what you don’t want to create new things. Effectively recycling on steroids. This can be manufacturing shoes, sustainable localized algae for plant growth, fashion, cars. Pretty much anything. This works and it truly is about self-sustainability. So how does it apply to AI? AI right now consumes everything, from data to power, to bandwidth, to server space, to data center real estate. Right now, it’s eating everything it can get its hands on. One would hope that eventually AI will figure out how to make itself more efficient. Today it’s not because we made it. Therefore, it’s inherently flawed and inefficient. When AI systems start remaking themselves, well that will be another problem for another day. This is almost a rebirth of that which has died, or perhaps a re-iteration of what will die? Where have a heard that one before…

Effective AI is going to start (and has already started) with autonomous machines. We’re there. We’ve arrived. Your cars already park themselves. Your cars can drive themselves. You’re already fearing which jobs are going to be lost to automation. Isn’t that what we were all afraid of during the industrial age, when Ford created factory conveyer belts and started building cars on a factory line? When bicycles were invented, the world went crazy without it was going to make everything obsolete. Does anyone ride a bicycle today? Few and far between, but proud to be one of them. Everything is factory line right now. We embraced it. We all have jobs and the world has evolved but not changed. So business isn’t going to change, but how we do business will change; its all about money. We stand at the feet of our oysters, and if there’s no money in it no one shucks.

The Business Landscape for Effective AI

As we explore the multifaceted world of Effective AI business models, it’s clear that the landscape is not one-dimensional; it encompasses diverse approaches and applications, each tailored to different needs and audiences.

Who will own the role of democratizing, decentralizing and securing AI for consumers and small-scale applications? This trial has undoubtedly made AI more accessible and intriguing. However, the true power of the machine is unleashed in the enterprise realm, where data-driven decision-making and automation drive significant efficiencies. Change the way you view the world, the way you view the business, and the way you view change.

The AI business landscape is a dynamic and evolving one, with opportunities and challenges aplenty. AI’s full potential is yet to be realized, and its journey is just beginning. Whether in the hands of individual users, small businesses, or large enterprises, AI’s impact is transformative and promises to shape the future of industries and economies. Let’s be smart and impactful.

Futurism at its best

My dream fantasy is to imagine yourself at a stop light (traffic light) and just sitting there for minutes that feel like hours. There’s no one on the road; it’s just you. No people. No cars. You wait and wait and wait, until finally you get that release when the light turns green. What was the point of you waiting like that? What if you were a pedestrian and there were no cars? Edge devices and IoT coupled with AI can change that experience.

Now re-imagine the idea of no cars, no people, and just you. Are you walking? Are you driving? Are you riding the infernal bicycle? There is no one else. The cameras detect what is there and more importantly what speed the cars and people are moving at, and then the lights change. Either you’re given the right of way to cross, or the right of way to keep driving through. No missed meetings, no late arrivals, no late dates, no accidents. No cares in the world. This is how we make smarter cities, smarter decisions, and smarter choices.

Think about it. This IoT scenario is a very real one. Smart cities will come, but we’re a bit of a way off. The infrastructure is here and we’re finally able to do it and handle the data needs. Remember 10,000 intersections in a city will consume data, and there are more than that in any average city. AI can help us manage these things. The situations that we don’t think about because we take them for granted.

This is The End, or Perhaps The Beginning

I’m just a person with a deep technology background, and I’ve delved into the challenges, and the pivotal role being played by AI in our current times. Where we are now isn’t our future. Our future will be bold and bright. The landscape of enterprise AI, its practical applications, and the steps required to realize its full potential. Simply put without a way to commercialize and make money, AI will remain a consumer grade toy for building art, for writing articles, for making press releases, and of course pushing fake everything; there is a politically destabilizing factor in here, where governments can take advantage of the technology and where they can own their statements. That is bold. Back in the day we seeded stories and ideas on the Internet and hoped that one day they would be relevant. I’ve been part of that landscape, but now the game has changed, but what do I know. I’m a futurist and a technologist. I just speak out and attempt to help people understand, and hopefully help you to understand what is real.

The business landscape for AI is a dynamic and multifaceted realm, with various players and approaches. While AI has made significant strides in making itself (That’s fun right? Makes it sound like it did it itself, but really, we did it) accessible to a broader audience, the heart of Effective AI’s power resides in its application in the enterprise sector. No revenue model, then nothing to do; it’s just marketing. I spoke to a friend the other day and he said that he spends 10% of his time on AI, there’s nothing there right now. I’m focused on enterprise revenue generation. If that doesn’t say it, then nothing does.

AI’s journey is far from over, and its impact is poised to shape industries, economies, politics, and the way we live and work. The quest for the perfect commercial model continues, as AI accelerates its influence across the globe, redefining what’s possible in the realm of technology and innovation can be felt. However, without a revenue model you don’t have anything to worry about. Your job is safe. So maybe let’s all take a breath and think about what tomorrow delivers that is in our grasp.

BETWEENPLAYS STOCKMARKET STRATEGIES

BETWEENPLAYS MEDIA

By Sandeep Panesar

Sandeep Panesar is a luminary in the realms of technology, entrepreneurship, and strategy. His journey began in the fields of telecommunications, media and entertainment, crafting innovative scripts and processes for the film and entertainment world. His tech background crosses telecommunications, edge security, cloud infrastructure, and emerging technologies. He spearheaded the growth of one of Canada's largest independent CLECs and carriers, leading eight acquisitions, and has forged global alliances with industry giants. He currently volunteers as a Governing Board representative and Chair for the Linux Foundation Networking Marketing council. Follow Sandeep @intosandeep on all social media platforms Instagram / X (twitter) / tik tok / ++ : @intosandeep Linkedin: https://www.linkedin.com/in/intosandeep IMDB: https://www.imdb.com/name/nm0997691