AI Operated Technology
AI Operated Technology
Exploring a future where AI takes technology, amplifying human creativity and innovation.
Welcome to the latest edition of ‘The API Economy’ — thanks for being here.
To support The API Economy, be like one of the 3,791 early, forward-thinking people who subscribe for monthly-ish insights on the API economy, Web3, AI, and more emerging trends.
Today's newsletter is brought to you by Ramp, the corporate card, bill pay, and reimbursement platform designed to help startups and venture funds save money with 1.5% cash-back on every swipe.
Ramp just raised $300m in fresh funding last week and is trusted by venture-backed startups from pre-seed through IPO like Morning Brew, Seed, Eventbrite, Discord, and Anduril.
AI has begun to transition from science fiction levels of expectation to the real-world architecture that will shape our digital lives. Its impact can be felt across sectors, introducing a paradigm shift that revolutionizes how we perceive, interact, and adapt to the world around us.
The promise and allure of AI lie in its ability to far surpass human capability.
But when it comes to operating the seemingly endless amounts of technology we have at our disposal today, the application of AI is akin to opening Pandora's box for humans. The process of operating modern software and technology stacks is now seeing the dawn of a new era where AI not merely complements but significantly outshines human efforts.
In the not-so-distant future, it will be clear. But who do you trust more to make decisions on behalf of technology? Humans or AI?
In theory, AI's ability to run software far exceeds that of humans, a proposition that might seem audacious, even threatening at first glance. But AI's true potential serves as an enabler, not a replacement for humans.
AI can liberate humans from mundane tasks and open the doors to uncharted territories of creativity and imagination.
The argument hinges on four essential aspects:
AI's aptitude for data-driven decision-making.
It has a considerably lower propensity for errors.
The diminishing need for human judgment in most software operations.
The liberation of human cognitive resources for tasks that require ingenuity and creativity.
When humans and AI coexist with their roles clearly defined, AI shoulders the burden of technical tasks, freeing humans to bask in the realm of creative endeavors.
Today, we’ll explore an exciting, challenging, but ultimately rewarding trajectory that charts the course of AI in software operation and the role of human intelligence in this brave new world.
Let’s dive in.
First chess, next human decision-making
AI's proficiency in data-driven decision-making already far exceeds the abilities humans could ever hope to possess. This aspect of AI, its ability to assimilate vast amounts of data, extract patterns, and base its decisions on this information, distinguishes it from its human counterparts.
Historically, chess was seen as a benchmark for human cognitive abilities. The game's complexity, with an estimated 10^120 possible positions, makes it an ideal platform for comparing human and artificial intelligence.
This comparison became a reality in 1997 when IBM's Deep Blue defeated World Chess Champion Garry Kasparov. Since then, AI's capabilities have only grown.
AI's success in chess is primarily due to its ability to evaluate millions of possible moves in seconds, far exceeding the processing speed of even the most skilled human players.
AI does this through an algorithmic approach called the Minimax algorithm, which predicts possible future moves and evaluates them based on their potential outcomes. AI is also free from cognitive biases or emotions, ensuring objective and consistent decision-making.
This is critical in chess because, by the 9th move, a player faces 2,439,530,234,167 possible games — an impossible number for the human brain to comprehend.
Another example demonstrating a paradigm shift in AI's approach to chess is how Google's DeepMind developed an AI program named AlphaZero. Compared to its predecessors, AlphaZero wasn't preprogrammed with any chess-specific knowledge. Instead, it learned the game from scratch using reinforcement learning, playing against itself millions of times and improving with each iteration.
Within a few hours of self-play, AlphaZero defeated Stockfish, one of the best chess engines in the world, in a 100-game match.
AlphaZero's triumph signified a significant leap in AI's data-driven decision-making capabilities. It showed us that given enough data and suitable learning algorithms, AI can teach itself complex tasks and outperform human experts. This example showcases the vast potential of AI in fields that are heavily dependent on pattern recognition, strategy, and decision-making, much like software operations.
The capabilities of AI in chess offer a window into the future, a future where AI could potentially surpass human decision-making in many more fields. It provides a compelling case for integrating AI in software operations, where decision-making speed and accuracy are paramount.
The chess example highlights what’s to come when AI is applied to more practical applications.
A McKinsey & Company study demonstrated that AI-powered systems could process and analyze data 100x faster than the average human analyst. AI's capacity for handling big data is almost limitless, an attribute that is becoming increasingly valuable in our data-driven world.
This prowess allows AI to make decisions based on comprehensive data analysis, eliminating the element of guesswork and improving the overall effectiveness and efficiency of software operations.
For example, Google's DeepMind AI was tasked with reducing the energy consumption of Google's data centers. Through real-time analysis and optimization, DeepMind achieved a 40% reduction in cooling costs, far exceeding any previously attempted human-led efforts.
While human decision-making is inherently influenced by emotions, cognitive biases, and individual experiences, AI operates purely on empirical data.
In a recent episode of the All-In Podcast, Chamath Palihapitiya said:
“AI is ruthless because it's emotionless. It was not taken to a steak dinner. It was not brought to a basketball game. It was not sold into a CEO. It's an agent that looked at a bunch of API endpoints, figured out how to write code to it to get done the job at hand.”
This objectivity results in decisions that are not only quicker but also more accurate and consistent. A report by Accenture found that decision-making errors by humans in software operations were 20% higher compared to AI.
The speed at which AI can make decisions provides another compelling advantage. According to an MIT Computer Science and Artificial Intelligence Laboratory study, AI can make complex data-driven decisions in milliseconds. In contrast, a human might take several minutes or even hours for the same task.
Despite the stellar capabilities of AI, it's essential to mention that AI's decisions are only as good as the data it is trained on. If the data is flawed, incomplete, or biased, the decisions made by AI will also be flawed. Therefore, rigorous data verification and regular updating of AI systems are necessary to ensure their continued effectiveness.
Through this exploration of AI's proficiency in data-driven decision-making, we begin to understand why AI outperforms humans in running software.
It sets the stage for our subsequent discussion on AI's lower propensity for errors, the reduced need for human judgment in software operations, and the freedom it offers humans to engage in more creative and innovative tasks.
In AI and The Burden of Knowledge, author Mario Gabriele states:
“As artificial intelligence accelerates past us, it will accumulate knowledge we cannot comprehend. Just as no amount of training can teach a dog calculus, these advanced models will be unable to explain their findings, thanks to our limited hardware. The result could be a modern life that humans increasingly do not understand, requiring a certain kind of faith.”
Anything you can do, AI can do better
A characteristic of AI that is often overlooked but profoundly important is the significantly lower propensity for errors than human operators.
The reality of human error is omnipresent across all fields and industries. Regardless of their experience and expertise, human operators are prone to making mistakes for many reasons, such as fatigue, cognitive biases, emotional states, or simply misinterpreting data.
According to a report by IBM, human error is responsible for approximately 95% of all security breaches in information technology.
In contrast, AI systems offer precision and consistency that is impossible to achieve by human standards. AI systems operate based on predefined algorithms and rules, free from the constraints of fatigue, emotion, or cognitive biases. They perform tasks the same way every time, ensuring high accuracy and precision.
One prime example of this accuracy is in the healthcare industry.
A study by Stanford University found that an AI algorithm designed to interpret medical images, such as X-rays and CT scans, could do so with an accuracy rate of 94%, outperforming human radiologists, who achieved an accuracy rate of 88%.
AI's precision is amplified when dealing with repetitive tasks, a mainstay in software operations. According to a study by McKinsey & Company, AI systems are up to five times more efficient at repetitive tasks than humans. This efficiency boosts productivity and significantly reduces the risk of errors, providing a compelling case for adopting AI in software operations.
These AI systems are also flexible. They are prone to errors if their algorithms are flawed or if the data they are trained on needs to be corrected or biased. Therefore, rigorous testing, regular updates, and continuous learning and improvement processes are crucial to maintaining the high accuracy rates of AI systems.
Human judgment just isn’t that good
Human judgment has long been the cornerstone of problem-solving and decision-making processes. People often refer to their decisions as “going with their gut.”
However, with the proliferation of AI and machine learning technologies, deciding to go with your gut might be a mistake. Thanks to AI, many tasks previously dependent on human judgment can now be automated or streamlined.
Nobel Prize-winning psychologist and co-author of the book “Noise: A Flaw in Human Judgment” Daniel Kahneman (also the author of the cult-classic “Thinking, Fast and Slow”) discussed the role of ‘noise’ in human judgment & how systematic decision-making can result in fewer errors in an interview with Science Friday saying:
“AI does better than reducing noise. Any algorithm, any systematic rule that takes inputs and combines them in a specified way, will have one crucial property – it will be noise-free. You present an algorithm with the same problem twice, you’re going to get the same answer.
But in general, algorithms are noise-free. And it turns out this is one of their major advantages over humans, that is, when you compare the performance of people to the performance of algorithms and rules, in many situations the algorithms and rules are already superior to people or match people. And the main reason for the lack of accuracy of people compared to algorithm is noise. People are noisy. Algorithms are not.”
A study conducted by the World Economic Forum found that by 2025, machines and algorithms will perform over 50% of all tasks in the workplace, up from 29% in 2018. This rise of automation is particularly noticeable in the software industry, where AI increasingly handles tasks like data analysis, debugging, and even coding.
Consider the example of automated testing in software development.
AI can identify patterns, predict possible defects, and execute tests much faster and more accurately than humans. According to a report by Capgemini, AI-powered testing can increase defect detection rates by up to 45% and reduce testing times by more than 50%.
AI models such as neural networks and deep learning algorithms can perform complex tasks, like interpreting natural language or recognizing patterns, which traditionally required human cognition. Google's BERT, a transformer-based model, is a case in point, understanding and generating human-like text at an unprecedented scale and efficiency.
However, the diminishing need for human judgment does not equate to its obsolescence. Instead, it underscores a shift in the roles humans need to play.
The complexities of AI and the ethical considerations surrounding its use necessitate a human touch. Therefore, instead of replacing human judgment, AI should be viewed as a tool that augments our capabilities, freeing us from mundane tasks and allowing us to focus on more creative and strategic aspects of our work.
AI for technology 🤝 Humans for creativity & innovation
The point of all this is not to feed the pervasive myth that the rise of AI and automation spells doom for human workers. Rather, the outcome of AI automating technical tasks for us is the liberation of human cognitive capacity for creative and innovative creation.
While it's true that AI can perform many tasks more efficiently and accurately than humans, it's also true that AI's role should be seen as complementary to human capabilities rather than as a replacement.
Consider a study by Accenture, which found that AI could increase business productivity by up to 40% by automating routine tasks. This automation frees up human time, allowing them to engage more deeply with creative, strategic, and socially engaging tasks - areas where AI still lags significantly behind human capabilities.
Introducing AI into software operations has profound implications for roles such as software developers and data analysts. Rather than spending time on mundane tasks like debugging code or cleaning data, they can focus on designing innovative algorithms, creating intuitive user interfaces, or discovering novel insights from data.
In a study by Adobe, 74% of business leaders reported that incorporating AI and machine learning into their operations gave their employees more time to be creative and strategic. As we integrate AI more deeply into our businesses and societies, new roles are emerging that blend technical knowledge with creativity and human understanding.
This transition is not automatic. To fully leverage the creative potential unleashed by AI, organizations must foster a culture of continuous learning and adaptability, invest in upskilling and reskilling their workforce, and redefine roles and workflows to encourage creativity and innovation.
AI <> human harmony
Picture a world where artificial intelligence and human ingenuity blend seamlessly, making one another better with each interaction. In this world, AI assumes the burden of routine tasks, processing data quickly and precisely far surpassing human capability.
In this world, AI gives us back the most valuable commodity in the world — time.
It is time to let our creativity blossom, innovate, and connect with each other on a deeper level. AI’s meticulous diligence allows us the freedom to harness the qualities that make us uniquely human: the ability to empathize, ideate, and inspire.
In this AI-enhanced utopia, we are not made obsolete but instead find our roles reimagined. We transition from being mere operators and overseers to creators and innovators. The AI revolution doesn't sideline us; it propels us into spheres where our strengths are best utilized.
On The Joe Rogan Postcast, Naval Ravikant discussed the automation revolutions, saying:
“This isn’t a new problem – throughout history, automation has replaced jobs. But it’s always freed people up for new creative work.
The question is not whether automation is going to eliminate jobs. We’ve been sitting around dividing up no finite number of jobs since the Stone Age. New jobs are being created, and are usually better and more creative. So the question is – how quickly is this transition going to happen, what kinds of jobs will be eliminated, and what kinds of jobs will be created?”
Consider the role of an artist or a poet, unburdened from mundane tasks, dedicating their newly freed time to creating a masterpiece or penning verses that touch souls. Visualize a researcher or a scientist, unhindered by data collection and processing, able to channel their efforts fully into innovative thinking and ground-breaking discoveries.
The future led by AI integration is not one of man versus machine but of man and machine. Together, in harmony, they advance towards a brighter future where technology serves not as a threat but as an enabler of human creativity and innovation.
It's a future that invites us not just to dream but to dream without limits. A world where AI is our tool, and the canvas of innovation is ours to paint.
Until our next adventure.
Special thanks to Mama Schroeder for editing this essay (any typos are on her 😊).
*Disclaimers:
The views in this essay are my own personal opinions and don’t necessarily represent the views of my employer, those mentioned in this article, or anyone other than myself.
Get $500 with Ramp after you spend your first $1K.
A modern economy circulating products and services throughout the world doesn’t need money or sovereign countries (national currencies) to be successful. Today, we’ve the scientific knowledge and technological skills to convert our natural and artificial resources into daily life-sustaining deliverables: food, housing, education, healthcare, infrastructure, and employment demands. What we lack is unity, a global framework built upon fair and humane laws and safe and healthy industrial practices. I hypothesize that humanity can end poverty and reduce pollution by abandoning wealth and property rights, and instead adopt and implement an advanced resource management system that can provide “universal protections for all”. Replacing customary political competition altogether, this type of approach, which I named facts-based representation, allows us a better way to govern ourselves and our communities, basing policy and decision making on the latest information, in turn improving the everyday outcomes impacting our personal and professional lives.
#ScientificSocialism