Stay Ahead, Stay ONMINE

😲 Quantifying Surprise – A Data Scientist’s Intro To Information Theory – Part 1/4: Foundations

Surprise! Generated using Gemini. During the telecommunication boom, Claude Shannon, in his seminal 1948 paper¹, posed a question that would revolutionise technology: How can we quantify communication? Shannon’s findings remain fundamental to expressing information quantification, storage, and communication. These insights made major contributions to the creation of technologies ranging from signal processing, data compression (e.g., Zip files and compact discs) to the Internet and artificial intelligence. More broadly, his work has significantly impacted diverse fields such as neurobiology, statistical physics and computer science (e.g, cybersecurity, cloud computing, and machine learning). [Shannon’s paper is the] Magna Carta of the Information Age Scientific American This is the first article in a series that explores information quantification – an essential tool for data scientists. Its applications range from enhancing statistical analyses to serving as a go-to decision heuristic in cutting-edge machine learning algorithms. Broadly speaking, quantifying information is assessing uncertainty, which may be phrased as: “how surprising is an outcome?”. This article idea quickly grew into a series since I found this topic both fascinating and diverse. Most researchers, at one stage or another, come across commonly used metrics such as entropy, cross-entropy/KL-divergence and mutual-information. Diving into this topic I found that in order to fully appreciate these one needs to learn a bit about the basics which we cover in this first article. By reading this series you will gain an intuition and tools to quantify: Bits/Nats – Unit measures of information. Self-Information – **** The amount of information in a specific event. Pointwise Mutual Information – The amount of information shared between two specific events. Entropy – The average amount of information of a variable’s outcome. Cross-entropy – The misalignment between two probability distributions (also expressed by its derivative KL-Divergence – a distance measure). Mutual Information – The co-dependency of two variables by their conditional probability distributions. It expresses the information gain of one variable given another. No prior knowledge is required – just a basic understanding of probabilities. I demonstrate using common statistics such as coin and dice 🎲 tosses as well as machine learning applications such as in supervised classification, feature selection, model monitoring and clustering assessment. As for real world applications I’ll discuss a case study of quantifying DNA diversity 🧬. Finally, for fun, I also apply to the popular brain twister commonly known as the Monty Hall problem 🚪🚪 🐐 . Throughout I provide python code 🐍 , and try to keep formulas as intuitive as possible. If you have access to an integrated development environment (IDE) 🖥 you might want to plug 🔌 and play 🕹 around with the numbers to gain a better intuition. This series is divided into four articles, each exploring a key aspect of Information Theory: 😲 Quantifying Surprise: 👈 👈 👈 YOU ARE HERE In this opening article, you’ll learn how to quantify the “surprise” of an event using _self-informatio_n and understand its units of measurement, such as _bit_s and _nat_s. Mastering self-information is essential for building intuition about the subsequent concepts, as all later heuristics are derived from it. 🤷 Quantifying Uncertainty: Building on self-information, this article shifts focus to the uncertainty – or “average surprise” – associated with a variable, known as entropy. We’ll dive into entropy’s wide-ranging applications, from Machine Learning and data analysis to solving fun puzzles, showcasing its adaptability. 📏 Quantifying Misalignment: Here, we’ll explore how to measure the distance between two probability distributions using entropy-based metrics like cross-entropy and KL-divergence. These measures are particularly valuable for tasks like comparing predicted versus true distributions, as in classification loss functions and other alignment-critical scenarios. 💸 Quantifying Gain: Expanding from single-variable measures, this article investigates the relationships between two. You’ll discover how to quantify the information gained about one variable (e.g, target Y) by knowing another (e.g., predictor X). Applications include assessing variable associations, feature selection, and evaluating clustering performance. Each article is crafted to stand alone while offering cross-references for deeper exploration. Together, they provide a practical, data-driven introduction to information theory, tailored for data scientists, analysts and machine learning practitioners. Disclaimer: Unless otherwise mentioned the formulas analysed are for categorical variables with c≥2 classes (2 meaning binary). Continuous variables will be addressed in a separate article. 🚧 Articles (3) and (4) are currently under construction. I will share links once available. Follow me to be notified 🚧 Quantifying Surprise with Self-Information Self-information is considered the building block of information quantification. It is a way of quantifying the amount of “surprise” of a specific outcome. Formally self-information, or also referred to as Shannon Information or information content, quantifies the surprise of an event x occurring based on its probability, p(x). Here we denote it as hₓ: Self-information _h_ₓ is the information of event x that occurs with probability p(x). The units of measure are called bits. One bit (binary digit) is the amount of information for an event x that has probability of p(x)=½. Let’s plug in to verify: hₓ=-log₂(½)= log₂(2)=1 bit. This heuristic serves as an alternative to probabilities, odds and log-odds, with certain mathematical properties which are advantageous for information theory. We discuss these below when learning about Shannon’s axioms behind this choice. It’s always informative to explore how an equation behaves with a graph: Bernoulli trial self-information h(p). Key features: Monotonic, h(p=1)=0, h(p →)→∞. To deepen our understanding of self-information, we’ll use this graph to explore the said axioms that justify its logarithmic formulation. Along the way, we’ll also build intuition about key features of this heuristic. To emphasise the logarithmic nature of self-information, I’ve highlighted three points of interest on the graph: At p=1 an event is guaranteed, yielding no surprise and hence zero bits of information (zero bits). A useful analogy is a trick coin (where both sides show HEAD). Reducing the probability by a factor of two (p=½​) increases the information to _hₓ=_1 bit. This, of course, is the case of a fair coin. Further reducing it by a factor of four results in hₓ(p=⅛)=3 bits. If you are interested in coding the graph here is a python script: To summarise this section: Self-Information hₓ=-log₂(p(x)) quantifies the amount of “surprise” of a specific outcome x. Three Axioms Referencing prior work by Ralph Hartley, Shannon chose -log₂(p) as a manner to meet three axioms. We’ll use the equation and graph to examine how these are manifested: An event with probability 100% is not surprising and hence does not yield any information. In the trick coin case this is evident by p(x)=1 yielding hₓ=0. Less probable events are more surprising and provide more information. This is apparent by self-information decreasing monotonically with increasing probability. The property of Additivity – the total self-information of two independent events equals the sum of individual contributions. This will be explored further in the upcoming fourth article on Mutual Information. There are mathematical proofs (which are beyond the scope of this series) that show that only the log function adheres to all three². The application of these axioms reveals several intriguing and practical properties of self-information: Important properties : Minimum bound: The first axiom hₓ(p=1)=0 establishes that self-information is non-negative, with zero as its lower bound. This is highly practical for many applications. Monotonically decreasing: The second axiom ensures that self-information decreases monotonically with increasing probability. No Maximum bound: At the extreme where _p→_0, monotonicity leads to self-information growing without bound hₓ(_p→0) →_ ∞, a feature that requires careful consideration in some contexts. However, when averaging self-information – as we will later see in the calculation of entropy – probabilities act as weights, effectively limiting the contribution of highly improbable events to the overall average. This relationship will become clearer when we explore entropy in detail. It is useful to understand the close relationship to log-odds. To do so we define p(x) as the probability of event x to happen and p(¬x)=1-p(x) of it not to happen. log-odds(x) = log₂(p(x)/p(¬x))= h(¬x) – h(x). The main takeaways from this section are Axiom 1: An event with probability 100% is not surprising Axiom 2: Less probable events are more surprising and, when they occur, provide more information. Self information (1) monotonically decreases (2) with a minimum bound of zero and (3) no upper bound. In the next two sections we further discuss units of measure and choice of normalisation. Information Units of Measure Bits or Shannons? A bit, as mentioned, represents the amount of information associated with an event that has a 50% probability of occurring. The term is also sometimes referred to as a Shannon, a naming convention proposed by mathematician and physicist David MacKay to avoid confusion with the term ‘bit’ in the context of digital processing and storage. After some deliberation, I decided to use ‘bit’ throughout this series for several reasons: This series focuses on quantifying information, not on digital processing or storage, so ambiguity is minimal. Shannon himself, encouraged by mathematician and statistician John Tukey, used the term ‘bit’ in his landmark paper. ‘Bit’ is the standard term in much of the literature on information theory. For convenience – it’s more concise Normalisation: Log Base 2 vs. Natural Throughout this series we use base 2 for logarithms, reflecting the intuitive notion of a 50% chance of an event as a fundamental unit of information. An alternative commonly used in machine learning is the natural logarithm, which introduces a different unit of measure called nats (short for natural units of information). One nat corresponds to the information gained from an event occurring with a probability of 1/e where e is Euler’s number (≈2.71828). In other words, 1 nat = -ln(p=(1/e)). The relationship between bits (base 2) and nats (natural log) is as follows: 1 bit = ln(2) nats ≈ 0.693 nats. Think of it as similar to a monetary current exchange or converting centimeters to inches. In his seminal publication Shanon explained that the optimal choice of base depends on the specific system being analysed (paraphrased slightly from his original work): “A device with two stable positions […] can store one bit of information” (bit as in binary digit). “A digit wheel on a desk computing machine that has ten stable positions […] has a storage capacity of one decimal digit.”³ “In analytical work where integration and differentiation are involved the base e is sometimes useful. The resulting units of information will be called natural units.” Key aspects of machine learning, such as popular loss functions, often rely on integrals and derivatives. The natural logarithm is a practical choice in these contexts because it can be derived and integrated without introducing additional constants. This likely explains why the machine learning community frequently uses nats as the unit of information – it simplifies the mathematics by avoiding the need to account for factors like ln(2). As shown earlier, I personally find base 2 more intuitive for interpretation. In cases where normalisation to another base is more convenient, I will make an effort to explain the reasoning behind the choice. To summarise this section of units of measure: bit = amount of information to distinguish between two equally likely outcomes. Now that we are familiar with self-information and its unit of measure let’s examine a few use cases. Quantifying Event Information with Coins and Dice In this section, we’ll explore examples to help internalise the self-information axioms and key features demonstrated in the graph. Gaining a solid understanding of self-information is essential for grasping its derivatives, such as entropy, cross-entropy (or KL divergence), and mutual information – all of which are averages over self-information. The examples are designed to be simple, approachable, and lighthearted, accompanied by practical Python code to help you experiment and build intuition. Note: If you feel comfortable with self-information, feel free to skip these examples and go straight to the Quantifying Uncertainty article. Generated using Gemini. To further explore the self-information and bits, I find analogies like coin flips and dice rolls particularly effective, as they are often useful analogies for real-world phenomena. Formally, these can be described as multinomial trials with n=1 trial. Specifically: A coin flip is a Bernoulli trial, where there are c=2 possible outcomes (e.g., heads or tails). Rolling a die represents a categorical trial, where c≥3 outcomes are possible (e.g., rolling a six-sided or eight-sided die). As a use case we’ll use simplistic weather reports limited to featuring sun 🌞 , rain 🌧 , and snow ⛄️. Now, let’s flip some virtual coins 👍 and roll some funky-looking dice 🎲 … Fair Coins and Dice Generated using Gemini. We’ll start with the simplest case of a fair coin (i.e, 50% chance for success/Heads or failure/Tails). Imagine an area for which at any given day there is a 50:50 chance for sun or rain. We can write the probability of each event be: p(🌞 )=p(🌧 )=½. As seen above, according the the self-information formulation, when 🌞 or 🌧 is reported we are provided with h(🌞 __ )=h(🌧 )=-log₂(½)=1 bit of information. We will continue to build on this analogy later on, but for now let’s turn to a variable that has more than two outcomes (c≥3). Before we address the standard six sided die, to simplify the maths and intuition, let’s assume an 8 sided one (_c=_8) as in Dungeons Dragons and other tabletop games. In this case each event (i.e, landing on each side) has a probability of p(🔲 ) = ⅛. When a die lands on one side facing up, e.g, value 7️⃣, we are provided with h(🔲 =7️⃣)=-log₂(⅛)=3 bits of information. For a standard six sided fair die: p(🔲 ) = ⅙ → an event yields __ h(🔲 )=-log₂(⅙)=2.58 bits. Comparing the amount of information from the fair coin (1 bit), 6 sided die (2.58 bits) and 8 sided (3 bits) we identify the second axiom: The less probable an event is, the more surprising it is and the more information it yields. Self information becomes even more interesting when probabilities are skewed to prefer certain events. Loaded Coins and Dice Generated using Gemini. Let’s assume a region where p(🌞 ) = ¾ and p(🌧 )= ¼. When rain is reported the amount of information conveyed is not 1 bit but rather h(🌧 )=-log₂(¼)=2 bits. When sun is reported less information is conveyed: h(🌞 )=-log₂(¾)=0.41 bits. As per the second axiom— a rarer event, like p(🌧 )=¼, reveals more information than a more likely one, like p(🌞 )=¾ – and vice versa. To further drive this point let’s now assume a desert region where p(🌞 ) =99% and p(🌧 )= 1%. If sunshine is reported – that is kind of expected – so nothing much is learnt (“nothing new under the sun” 🥁) and this is quantified as h(🌞 )=0.01 bits. If rain is reported, however, you can imagine being quite surprised. This is quantified as h(🌧 )=6.64 bits. In the following python scripts you can examine all the above examples, and I encourage you to play with your own to get a feeling. First let’s define the calculation and printout function: import numpy as np def print_events_self_information(probs): for ps in probs: print(f”Given distribution {ps}”) for event in ps: if ps[event] != 0: self_information = -np.log2(ps[event]) #same as: -np.log(ps[event])/np.log(2) text_ = f’When `{event}` occurs {self_information:0.2f} bits of information is communicated’ print(text_) else: print(f’a `{event}` event cannot happen p=0 ‘) print(“=” * 20) Next we’ll set a few example distributions of weather frequencies # Setting multiple probability distributions (each sums to 100%) # Fun fact – 🐍 💚 Emojis! probs = [{‘🌞 ‘: 0.5, ‘🌧 ‘: 0.5}, # half-half {‘🌞 ‘: 0.75, ‘🌧 ‘: 0.25}, # more sun than rain {‘🌞 ‘: 0.99, ‘🌧 ‘: 0.01} , # mostly sunshine ] print_events_self_information(probs) This yields printout Given distribution {‘🌞 ‘: 0.5, ‘🌧 ‘: 0.5} When `🌞 ` occurs 1.00 bits of information is communicated When `🌧 ` occurs 1.00 bits of information is communicated ==================== Given distribution {‘🌞 ‘: 0.75, ‘🌧 ‘: 0.25} When `🌞 ` occurs 0.42 bits of information is communicated When `🌧 ` occurs 2.00 bits of information is communicated ==================== Given distribution {‘🌞 ‘: 0.99, ‘🌧 ‘: 0.01} When `🌞 ` occurs 0.01 bits of information is communicated When `🌧 ` occurs 6.64 bits of information is communicated Let’s examine a case of a loaded three sided die. E.g, information of a weather in an area that reports sun, rain and snow at uneven probabilities: p(🌞 ) = 0.2, p(🌧 )=0.7, p(⛄️)=0.1. Running the following print_events_self_information([{‘🌞 ‘: 0.2, ‘🌧 ‘: 0.7, ‘⛄️’: 0.1}]) yields Given distribution {‘🌞 ‘: 0.2, ‘🌧 ‘: 0.7, ‘⛄️’: 0.1} When `🌞 ` occurs 2.32 bits of information is communicated When `🌧 ` occurs 0.51 bits of information is communicated When `⛄️` occurs 3.32 bits of information is communicated What we saw for the binary case applies to higher dimensions. To summarise – we clearly see the implications of the second axiom: When a highly expected event occurs – we do not learn much, the bit count is low. When an unexpected event occurs – we learn a lot, the bit count is high. Event Information Summary In this article we embarked on a journey into the foundational concepts of information theory, defining how to measure the surprise of an event. Notions introduced serve as the bedrock of many tools in information theory, from assessing data distributions to unraveling the inner workings of machine learning algorithms. Through simple yet insightful examples like coin flips and dice rolls, we explored how self-information quantifies the unpredictability of specific outcomes. Expressed in bits, this measure encapsulates Shannon’s second axiom: rarer events convey more information. While we’ve focused on the information content of specific events, this naturally leads to a broader question: what is the average amount of information associated with all possible outcomes of a variable? In the next article, Quantifying Uncertainty, we build on the foundation of self-information and bits to explore entropy – the measure of average uncertainty. Far from being just a beautiful theoretical construct, it has practical applications in data analysis and machine learning, powering tasks like decision tree optimisation, estimating diversity and more. Claude Shannon. Credit: Wikipedia Loved this post? ❤️🍕 💌 Follow me here, join me on LinkedIn or 🍕 buy me a pizza slice! About This Series Even though I have twenty years of experience in data analysis and predictive modelling I always felt quite uneasy about using concepts in information theory without truly understanding them. The purpose of this series was to put me more at ease with concepts of information theory and hopefully provide for others the explanations I needed. 🤷 Quantifying Uncertainty – A Data Scientist’s Intro To Information Theory – Part 2/4: EntropyGa_in intuition into Entropy and master its applications in Machine Learning and Data Analysis. Python code included. 🐍 me_dium.com Check out my other articles which I wrote to better understand Causality and Bayesian Statistics: Footnotes ¹ A Mathematical Theory of Communication, Claude E. Shannon, Bell System Technical Journal 1948. It was later renamed to a book The Mathematical Theory of Communication in 1949. [Shannon’s “A Mathematical Theory of Communication”] the blueprint for the digital era – Historian James Gleick ² See Wikipedia page on Information Content (i.e, self-information) for a detailed derivation that only the log function meets all three axioms. ³ The decimal-digit was later renamed to a hartley (symbol Hart), a ban or a dit. See Hartley (unit) Wikipedia page. Credits Unless otherwise noted, all images were created by the author. Many thanks to Will Reynolds and Pascal Bugnion for their useful comments.
Surprise! Generated using Gemini.
Surprise! Generated using Gemini.

During the telecommunication boom, Claude Shannon, in his seminal 1948 paper¹, posed a question that would revolutionise technology:

How can we quantify communication?

Shannon’s findings remain fundamental to expressing information quantification, storage, and communication. These insights made major contributions to the creation of technologies ranging from signal processing, data compression (e.g., Zip files and compact discs) to the Internet and artificial intelligence. More broadly, his work has significantly impacted diverse fields such as neurobiology, statistical physics and computer science (e.g, cybersecurity, cloud computing, and machine learning).

[Shannon’s paper is the]

Magna Carta of the Information Age

  • Scientific American

This is the first article in a series that explores information quantification – an essential tool for data scientists. Its applications range from enhancing statistical analyses to serving as a go-to decision heuristic in cutting-edge machine learning algorithms.

Broadly speaking, quantifying information is assessing uncertainty, which may be phrased as: “how surprising is an outcome?”.

This article idea quickly grew into a series since I found this topic both fascinating and diverse. Most researchers, at one stage or another, come across commonly used metrics such as entropy, cross-entropy/KL-divergence and mutual-information. Diving into this topic I found that in order to fully appreciate these one needs to learn a bit about the basics which we cover in this first article.

By reading this series you will gain an intuition and tools to quantify:

  • Bits/Nats – Unit measures of information.
  • Self-Information – **** The amount of information in a specific event.
  • Pointwise Mutual Information – The amount of information shared between two specific events.
  • Entropy – The average amount of information of a variable’s outcome.
  • Cross-entropy – The misalignment between two probability distributions (also expressed by its derivative KL-Divergence – a distance measure).
  • Mutual Information – The co-dependency of two variables by their conditional probability distributions. It expresses the information gain of one variable given another.

No prior knowledge is required – just a basic understanding of probabilities.

I demonstrate using common statistics such as coin and dice 🎲 tosses as well as machine learning applications such as in supervised classification, feature selection, model monitoring and clustering assessment. As for real world applications I’ll discuss a case study of quantifying DNA diversity 🧬. Finally, for fun, I also apply to the popular brain twister commonly known as the Monty Hall problem 🚪🚪 🐐 .

Throughout I provide python code 🐍 , and try to keep formulas as intuitive as possible. If you have access to an integrated development environment (IDE) 🖥 you might want to plug 🔌 and play 🕹 around with the numbers to gain a better intuition.

This series is divided into four articles, each exploring a key aspect of Information Theory:

  1. 😲 Quantifying Surprise: 👈 👈 👈 YOU ARE HERE
    In this opening article, you’ll learn how to quantify the “surprise” of an event using _self-informatio_n and understand its units of measurement, such as _bit_s and _nat_s. Mastering self-information is essential for building intuition about the subsequent concepts, as all later heuristics are derived from it.

  2. 🤷 Quantifying Uncertainty: Building on self-information, this article shifts focus to the uncertainty – or “average surprise” – associated with a variable, known as entropy. We’ll dive into entropy’s wide-ranging applications, from Machine Learning and data analysis to solving fun puzzles, showcasing its adaptability.
  3. 📏 Quantifying Misalignment: Here, we’ll explore how to measure the distance between two probability distributions using entropy-based metrics like cross-entropy and KL-divergence. These measures are particularly valuable for tasks like comparing predicted versus true distributions, as in classification loss functions and other alignment-critical scenarios.
  4. 💸 Quantifying Gain: Expanding from single-variable measures, this article investigates the relationships between two. You’ll discover how to quantify the information gained about one variable (e.g, target Y) by knowing another (e.g., predictor X). Applications include assessing variable associations, feature selection, and evaluating clustering performance.

Each article is crafted to stand alone while offering cross-references for deeper exploration. Together, they provide a practical, data-driven introduction to information theory, tailored for data scientists, analysts and machine learning practitioners.

Disclaimer: Unless otherwise mentioned the formulas analysed are for categorical variables with c≥2 classes (2 meaning binary). Continuous variables will be addressed in a separate article.

🚧 Articles (3) and (4) are currently under construction. I will share links once available. Follow me to be notified 🚧


Quantifying Surprise with Self-Information

Self-information is considered the building block of information quantification.

It is a way of quantifying the amount of “surprise” of a specific outcome.

Formally self-information, or also referred to as Shannon Information or information content, quantifies the surprise of an event x occurring based on its probability, p(x). Here we denote it as hₓ:

Self-information _h_ₓ is the information of event x that occurs with probability p(x).
Self-information _h_ₓ is the information of event x that occurs with probability p(x).

The units of measure are called bits. One bit (binary digit) is the amount of information for an event x that has probability of p(x)=½. Let’s plug in to verify: hₓ=-log₂(½)= log₂(2)=1 bit.

This heuristic serves as an alternative to probabilities, odds and log-odds, with certain mathematical properties which are advantageous for information theory. We discuss these below when learning about Shannon’s axioms behind this choice.

It’s always informative to explore how an equation behaves with a graph:

Bernoulli trial self-information h(p). Key features: Monotonic, h(p=1)=0, h(p →)→∞.
Bernoulli trial self-information h(p). Key features: Monotonic, h(p=1)=0, h(p →)→∞.

To deepen our understanding of self-information, we’ll use this graph to explore the said axioms that justify its logarithmic formulation. Along the way, we’ll also build intuition about key features of this heuristic.

To emphasise the logarithmic nature of self-information, I’ve highlighted three points of interest on the graph:

  • At p=1 an event is guaranteed, yielding no surprise and hence zero bits of information (zero bits). A useful analogy is a trick coin (where both sides show HEAD).
  • Reducing the probability by a factor of two (p=½​) increases the information to _hₓ=_1 bit. This, of course, is the case of a fair coin.
  • Further reducing it by a factor of four results in hₓ(p=⅛)=3 bits.

If you are interested in coding the graph here is a python script:

To summarise this section:

Self-Information hₓ=-log₂(p(x)) quantifies the amount of “surprise” of a specific outcome x.

Three Axioms

Referencing prior work by Ralph Hartley, Shannon chose -log₂(p) as a manner to meet three axioms. We’ll use the equation and graph to examine how these are manifested:

  1. An event with probability 100% is not surprising and hence does not yield any information.
    In the trick coin case this is evident by p(x)=1 yielding hₓ=0.

  2. Less probable events are more surprising and provide more information.
    This is apparent by self-information decreasing monotonically with increasing probability.

  3. The property of Additivity – the total self-information of two independent events equals the sum of individual contributions. This will be explored further in the upcoming fourth article on Mutual Information.

There are mathematical proofs (which are beyond the scope of this series) that show that only the log function adheres to all three².

The application of these axioms reveals several intriguing and practical properties of self-information:

Important properties :

  • Minimum bound: The first axiom hₓ(p=1)=0 establishes that self-information is non-negative, with zero as its lower bound. This is highly practical for many applications.
  • Monotonically decreasing: The second axiom ensures that self-information decreases monotonically with increasing probability.
  • No Maximum bound: At the extreme where _p→_0, monotonicity leads to self-information growing without bound hₓ(_p→0) →_ ∞, a feature that requires careful consideration in some contexts. However, when averaging self-information – as we will later see in the calculation of entropy – probabilities act as weights, effectively limiting the contribution of highly improbable events to the overall average. This relationship will become clearer when we explore entropy in detail.

It is useful to understand the close relationship to log-odds. To do so we define p(x) as the probability of event x to happen and p(¬x)=1-p(x) of it not to happen. log-odds(x) = log₂(p(x)/p(¬x))= h(¬x) – h(x).

The main takeaways from this section are

Axiom 1: An event with probability 100% is not surprising

Axiom 2: Less probable events are more surprising and, when they occur, provide more information.

Self information (1) monotonically decreases (2) with a minimum bound of zero and (3) no upper bound.

In the next two sections we further discuss units of measure and choice of normalisation.

Information Units of Measure

Bits or Shannons?

A bit, as mentioned, represents the amount of information associated with an event that has a 50% probability of occurring.

The term is also sometimes referred to as a Shannon, a naming convention proposed by mathematician and physicist David MacKay to avoid confusion with the term ‘bit’ in the context of digital processing and storage.

After some deliberation, I decided to use ‘bit’ throughout this series for several reasons:

  • This series focuses on quantifying information, not on digital processing or storage, so ambiguity is minimal.
  • Shannon himself, encouraged by mathematician and statistician John Tukey, used the term ‘bit’ in his landmark paper.
  • ‘Bit’ is the standard term in much of the literature on information theory.
  • For convenience – it’s more concise

Normalisation: Log Base 2 vs. Natural

Throughout this series we use base 2 for logarithms, reflecting the intuitive notion of a 50% chance of an event as a fundamental unit of information.

An alternative commonly used in machine learning is the natural logarithm, which introduces a different unit of measure called nats (short for natural units of information). One nat corresponds to the information gained from an event occurring with a probability of 1/e where e is Euler’s number (≈2.71828). In other words, 1 nat = -ln(p=(1/e)).

The relationship between bits (base 2) and nats (natural log) is as follows:

1 bit = ln(2) nats ≈ 0.693 nats.

Think of it as similar to a monetary current exchange or converting centimeters to inches.

In his seminal publication Shanon explained that the optimal choice of base depends on the specific system being analysed (paraphrased slightly from his original work):

  • “A device with two stable positions […] can store one bit of information” (bit as in binary digit).
  • “A digit wheel on a desk computing machine that has ten stable positions […] has a storage capacity of one decimal digit.”³
  • “In analytical work where integration and differentiation are involved the base e is sometimes useful. The resulting units of information will be called natural units.“

Key aspects of machine learning, such as popular loss functions, often rely on integrals and derivatives. The natural logarithm is a practical choice in these contexts because it can be derived and integrated without introducing additional constants. This likely explains why the machine learning community frequently uses nats as the unit of information – it simplifies the mathematics by avoiding the need to account for factors like ln(2).

As shown earlier, I personally find base 2 more intuitive for interpretation. In cases where normalisation to another base is more convenient, I will make an effort to explain the reasoning behind the choice.

To summarise this section of units of measure:

bit = amount of information to distinguish between two equally likely outcomes.

Now that we are familiar with self-information and its unit of measure let’s examine a few use cases.

Quantifying Event Information with Coins and Dice

In this section, we’ll explore examples to help internalise the self-information axioms and key features demonstrated in the graph. Gaining a solid understanding of self-information is essential for grasping its derivatives, such as entropy, cross-entropy (or KL divergence), and mutual information – all of which are averages over self-information.

The examples are designed to be simple, approachable, and lighthearted, accompanied by practical Python code to help you experiment and build intuition.

Note: If you feel comfortable with self-information, feel free to skip these examples and go straight to the Quantifying Uncertainty article.

Generated using Gemini.
Generated using Gemini.

To further explore the self-information and bits, I find analogies like coin flips and dice rolls particularly effective, as they are often useful analogies for real-world phenomena. Formally, these can be described as multinomial trials with n=1 trial. Specifically:

  • A coin flip is a Bernoulli trial, where there are c=2 possible outcomes (e.g., heads or tails).
  • Rolling a die represents a categorical trial, where c≥3 outcomes are possible (e.g., rolling a six-sided or eight-sided die).

As a use case we’ll use simplistic weather reports limited to featuring sun 🌞 , rain 🌧 , and snow ⛄️.

Now, let’s flip some virtual coins 👍 and roll some funky-looking dice 🎲 …

Fair Coins and Dice

Generated using Gemini.
Generated using Gemini.

We’ll start with the simplest case of a fair coin (i.e, 50% chance for success/Heads or failure/Tails).

Imagine an area for which at any given day there is a 50:50 chance for sun or rain. We can write the probability of each event be: p(🌞 )=p(🌧 )=½.

As seen above, according the the self-information formulation, when 🌞 or 🌧 is reported we are provided with h(🌞 __ )=h(🌧 )=-log₂(½)=1 bit of information.

We will continue to build on this analogy later on, but for now let’s turn to a variable that has more than two outcomes (c≥3).

Before we address the standard six sided die, to simplify the maths and intuition, let’s assume an 8 sided one (_c=_8) as in Dungeons Dragons and other tabletop games. In this case each event (i.e, landing on each side) has a probability of p(🔲 ) = ⅛.

When a die lands on one side facing up, e.g, value 7️⃣, we are provided with h(🔲 =7️⃣)=-log₂(⅛)=3 bits of information.

For a standard six sided fair die: p(🔲 ) = ⅙ → an event yields __ h(🔲 )=-log₂(⅙)=2.58 bits.

Comparing the amount of information from the fair coin (1 bit), 6 sided die (2.58 bits) and 8 sided (3 bits) we identify the second axiom: The less probable an event is, the more surprising it is and the more information it yields.

Self information becomes even more interesting when probabilities are skewed to prefer certain events.

Loaded Coins and Dice

Generated using Gemini.
Generated using Gemini.

Let’s assume a region where p(🌞 ) = ¾ and p(🌧 )= ¼.

When rain is reported the amount of information conveyed is not 1 bit but rather h(🌧 )=-log₂(¼)=2 bits.

When sun is reported less information is conveyed: h(🌞 )=-log₂(¾)=0.41 bits.

As per the second axiom— a rarer event, like p(🌧 )=¼, reveals more information than a more likely one, like p(🌞 )=¾ – and vice versa.

To further drive this point let’s now assume a desert region where p(🌞 ) =99% and p(🌧 )= 1%.

If sunshine is reported – that is kind of expected – so nothing much is learnt (“nothing new under the sun” 🥁) and this is quantified as h(🌞 )=0.01 bits. If rain is reported, however, you can imagine being quite surprised. This is quantified as h(🌧 )=6.64 bits.

In the following python scripts you can examine all the above examples, and I encourage you to play with your own to get a feeling.

First let’s define the calculation and printout function:

import numpy as np

def print_events_self_information(probs):
    for ps in probs:
        print(f"Given distribution {ps}")
        for event in ps:
            if ps[event] != 0:
                self_information = -np.log2(ps[event]) #same as: -np.log(ps[event])/np.log(2) 
                text_ = f'When `{event}` occurs {self_information:0.2f} bits of information is communicated'
                print(text_)
            else:
                print(f'a `{event}` event cannot happen p=0 ')
        print("=" * 20)

Next we’ll set a few example distributions of weather frequencies

# Setting multiple probability distributions (each sums to 100%)
# Fun fact - 🐍  💚  Emojis!
probs = [{'🌞   ': 0.5, '🌧   ': 0.5},   # half-half
        {'🌞   ': 0.75, '🌧   ': 0.25},  # more sun than rain
        {'🌞   ': 0.99, '🌧   ': 0.01} , # mostly sunshine
]

print_events_self_information(probs)

This yields printout

Given distribution {'🌞      ': 0.5, '🌧      ': 0.5}
When `🌞      ` occurs 1.00 bits of information is communicated 
When `🌧      ` occurs 1.00 bits of information is communicated 
====================
Given distribution {'🌞      ': 0.75, '🌧      ': 0.25}
When `🌞      ` occurs 0.42 bits of information is communicated 
When `🌧      ` occurs 2.00 bits of information is communicated 
====================
Given distribution {'🌞      ': 0.99, '🌧      ': 0.01}
When `🌞      ` occurs 0.01 bits of information is communicated 
When `🌧      ` occurs 6.64 bits of information is communicated  

Let’s examine a case of a loaded three sided die. E.g, information of a weather in an area that reports sun, rain and snow at uneven probabilities: p(🌞 ) = 0.2, p(🌧 )=0.7, p(⛄️)=0.1.

Running the following

print_events_self_information([{'🌞 ': 0.2, '🌧 ': 0.7, '⛄️': 0.1}])

yields

Given distribution {'🌞  ': 0.2, '🌧  ': 0.7, '⛄️': 0.1}
When `🌞  ` occurs 2.32 bits of information is communicated 
When `🌧  ` occurs 0.51 bits of information is communicated 
When `⛄️` occurs 3.32 bits of information is communicated 

What we saw for the binary case applies to higher dimensions.

To summarise – we clearly see the implications of the second axiom:

  • When a highly expected event occurs – we do not learn much, the bit count is low.
  • When an unexpected event occurs – we learn a lot, the bit count is high.

Event Information Summary

In this article we embarked on a journey into the foundational concepts of information theory, defining how to measure the surprise of an event. Notions introduced serve as the bedrock of many tools in information theory, from assessing data distributions to unraveling the inner workings of machine learning algorithms.

Through simple yet insightful examples like coin flips and dice rolls, we explored how self-information quantifies the unpredictability of specific outcomes. Expressed in bits, this measure encapsulates Shannon’s second axiom: rarer events convey more information.

While we’ve focused on the information content of specific events, this naturally leads to a broader question: what is the average amount of information associated with all possible outcomes of a variable?

In the next article, Quantifying Uncertainty, we build on the foundation of self-information and bits to explore entropy – the measure of average uncertainty. Far from being just a beautiful theoretical construct, it has practical applications in data analysis and machine learning, powering tasks like decision tree optimisation, estimating diversity and more.

Claude Shannon. Credit: Wikipedia
Claude Shannon. Credit: Wikipedia

Loved this post? ❤️🍕

💌 Follow me here, join me on LinkedIn or 🍕 buy me a pizza slice!

About This Series

Even though I have twenty years of experience in data analysis and predictive modelling I always felt quite uneasy about using concepts in information theory without truly understanding them.

The purpose of this series was to put me more at ease with concepts of information theory and hopefully provide for others the explanations I needed.

🤷 Quantifying Uncertainty – A Data Scientist’s Intro To Information Theory – Part 2/4: EntropyGa_in intuition into Entropy and master its applications in Machine Learning and Data Analysis. Python code included. 🐍 me_dium.com

Check out my other articles which I wrote to better understand Causality and Bayesian Statistics:

Footnotes

¹ A Mathematical Theory of Communication, Claude E. Shannon, Bell System Technical Journal 1948.

It was later renamed to a book The Mathematical Theory of Communication in 1949.

[Shannon’s “A Mathematical Theory of Communication”] the blueprint for the digital era – Historian James Gleick

² See Wikipedia page on Information Content (i.e, self-information) for a detailed derivation that only the log function meets all three axioms.

³ The decimal-digit was later renamed to a hartley (symbol Hart), a ban or a dit. See Hartley (unit) Wikipedia page.

Credits

Unless otherwise noted, all images were created by the author.

Many thanks to Will Reynolds and Pascal Bugnion for their useful comments.

Shape
Shape
Stay Ahead

Explore More Insights

Stay ahead with more perspectives on cutting-edge power, infrastructure, energy,  bitcoin and AI solutions. Explore these articles to uncover strategies and insights shaping the future of industries.

Shape

What’s next for Nvidia?

That’s if he wants it. One industry watcher told this reporter that several years ago, he asked several Nvidia vice presidents if they wanted the big chair, because Nvidia had no clear successor to Huang and to a man they said no way. “I can’t and don’t want to do

Read More »

AMD takes on Intel and Nvidia with new 256-core EPYC server CPUs

AMD EPYC, Intel Xeon, and Nvidia Vera represent three different approaches to the evolving data-centre CPU market, added Rawat. “AMD competes through high core density, memory capabilities and power efficiency, while Intel leverages its established enterprise ecosystem, platform compatibility and customer relationships. Nvidia is taking a different approach with Vera,

Read More »

Energy Secretary Keeps Critical, Coal-Fired Generation Operational Ahead of Winter in the Midwest

WASHINGTON—The U.S. Department of Energy (DOE) issued an emergency order to address critical grid reliability issues in the Midwest. The emergency order directs the Midcontinent Independent System Operator (MISO), in coordination with Consumers Energy Company, to ensure that the coal-fired J.H. Campbell Generating Plant (Campbell Plant) in West Olive, Michigan, is available to operate and meet emergency needs while employing economic dispatch to minimize costs for hardworking American families and businesses. The Campbell Plant was originally slated to shut down on May 31, 2025, 15 years before the end of its scheduled design life.  “President Trump and the Energy Department remain committed to doing everything in our power to stop power outages for American families and businesses,” said U.S. Secretary of Energy Chris Wright. “Ensuring coal plants such as the Campbell Plant are available to operate has prevented blackouts and likely saved lives. Americans deserve access to affordable, reliable, and secure electricity regardless of whether the wind is blowing or the sun is shining.” Thanks to President Trump, coal generating plants across the country are being saved from premature retirement. For example, in 2025, more than 17 gigawatts of coal-fired electricity generation were saved from going offline. During periods of peak electricity demand, many of those plants keep the lights on.  The emergency conditions that led to the issuance of the original order persist. A joint report from the Federal Energy Regulatory Commission (FERC), the North American Electric Reliability Corporation (NERC), and NERC’s Regional Entities highlights how events that occurred during the Winter 2026 Arctic Cold Period, including Winter Storm Fern, demonstrated the need for dispatchable generation capacity from facilities like the Campbell Plant.  This order is in effect from November 15, 2026, through February 12, 2027. 

Read More »

Kosmos lets contract to advance Tiberius project offshore Louisiana

Kosmos Energy Gulf of Mexico Operations LLC, a unit of Kosmos Energy, has let a contract to Subsea 7 SA for work on the Tiberius project in the US Gulf of Mexico. The scope of work covers the fabrication, transportation, and installation of a 14-km steel catenary riser and wet-insulated flowline to the Lucius Spar, along with the installation of an umbilical and subsea controls. The Tiberius project lies in Keathley Canyon lease 964 about 400 km south of New Orleans in about 2,283 m of water.  License partners reached a final investment decision on the Phase 1A development this year, which includes completion of an already drilled well. Long-lead items have been secured, and first oil is expected by the end of third-quarter 2028.  Subsea 7, which values the contract at $50-150 million, expects project management and engineering activities to begin in the coming months, with offshore activities expected to begin in 2028. Kosmos is operator of the Tiberius project with a 33.34% interest following a farmout alongside partner Occidental Petroleum to Navitas Petroleum US LLC earlier this year.

Read More »

ExxonMobil takes FID for Longtail development offshore Guyana

The development, in the southeastern portion of the block, will target non-associated gas and condensate-rich reservoirs. Subject to future investment decisions, market demand, and development of supporting infrastructure, gas from the project could be transported to shore by pipeline. The development will use a floating production, storage, and offloading (FPSO) vessel designed to process 1.2 bcfd of gas and produce as much as 250,000 b/d of condensate. The project is expected to include about 45 subsea production and injection wells. “By developing these gas and condensate resources, we’re unlocking new value for Guyana while delivering advantaged energy to meet growing global demand,” said Dan Ammann, president, ExxonMobil Upstream Company. ExxonMobil currently produces about 900,000 b/d of oil from four developments in the Stabroek block. Its fifth project, Uaru, is expected to start up in fourth-quarter 2026, increasing total block production capacity to more than 1 million b/d. Earlier this year, the operator and co-venturers produced the 1 billionth bbl of oil from Guyana’s Stabroek block, less than 7 years after first production began in December 2019 with Liza field. Construction also continues on the Whiptail and Hammerhead developments, the sixth and seventh approved projects on the block. Whiptail is slated to start production in 2027, followed by Hammerhead in 2029. ExxonMobil Guyana Ltd. operates the Stabroek block with a 45% interest. Hess Guyana Exploration Ltd. holds 30%, and CNOOC Petroleum Guyana Ltd. holds 25%.

Read More »

MARAD approves Energy Transfer’s $1-billion Blue Marlin crude oil export port

The US Maritime Administration (MARAD) approved Energy Transfer’s $1-billion Blue Marlin offshore crude oil export project off Louisiana Nov. 5, marking the fourth oil and gas deepwater port approval under President Donald Trump’s second administration. The proposed Blue Marlin port would have capacity to export up to 2 million b/d of crude oil, with loading rates of up to 80,000 bbl/hr, including onto very large crude carriers (VLCCs). Located 99 miles off Cameron Parish, La., the project includes conversion of an existing 36-in. Stingray natural gas pipeline to transport crude oil from Energy Transfer’s Nederland terminal in Texas to the offshore infrastructure. The US Federal Energy Regulatory Commission previously authorized abandonment and sale of Stingray pipeline infrastructure for conversion to crude oil service. Energy Transfer has yet to receive final US Environmental Protection Agency air permits and certain state approvals. The project’s commercial development schedule remains uncertain. MARAD approved Blue Marlin shortly after its Oct. 28 approval of the $8.2-billion ST LNG project offshore Texas, as the agency pursues new procedures to streamline deepwater port licensing. The Trump administration also has approved Delfin LNG, a floating LNG export project offshore Louisiana, and Texas GulfLink, a crude oil export project offshore Texas.

Read More »

IEA sees oil market surplus returning as Gulf exports recover

The global oil market is showing signs of rebalancing as recovering crude exports from the Middle East lift supply, although persistent disruptions to refined product markets, particularly diesel, continue to weigh on consumers, the International Energy Agency (IEA) said. In its Nov. 13 Oil Market Report, IEA said global oil supply increased 2.3 million b/d month-on-month to 105.4 million b/d in October, following a 3 million-b/d increase in September. Output has risen 5.3 million b/d since August as Gulf exports recovered despite continued attacks on energy infrastructure and commercial tankers. The stronger supply outlook has prompted the agency to raise its fourth-quarter 2026 supply estimate by 3.3 million b/d. Global supply is now expected to decline by an average 4.8 million b/d for the full year before rebounding by 6.6 million b/d in 2027 to average 108.1 million b/d. The recovery in exports has eased market tightness and pushed crude prices lower. North Sea Dated crude fell 32% from an early-October high of $136/bbl to about $93/bbl at the time of the report, although prices remained $22/bbl above prewar levels. Gulf oil exports, excluding Oman, reached 18.4 million b/d in October, about 26% below prewar levels. Crude exports accounted for most of the recovery, reaching 15 million b/d, equivalent to 82% of prewar volumes. However, IEA cautioned that continued attacks on shipping remain a significant downside risk to the supply outlook. Demand under pressure Global oil demand is forecast to decline by 2.7 million b/d in 2026 to 102.3 million b/d, reflecting disruptions to refined product and liquefied petroleum gas exports from the Gulf, elevated fuel prices, and weaker consumption across major markets. The contraction is expected to moderate in fourth-quarter 2026 to 2.7 million b/d year-on-year, compared with an average decline of 4.4 million b/d in the second and

Read More »

What’s next for California’s climate laws following an appeals court split decision

The Ninth Circuit Court of Appeals issued a split decision on California’s climate disclosure laws, Senate Bill 253 and Senate Bill 261, on Monday, allowing some emissions reporting requirements to proceed, while blocking others on the basis of them being considered in part or completely unconstitutional under the First Amendment. In the decision, the federal appellate panel partially upheld and partially overturned a district court’s decision from August 2025, which declined to enjoin the laws. The U.S. Chamber of Commerce and other business associations filed the lawsuit, requesting to temporarily block implementation of both climate laws while their constitutionality was under review. Implementation of SB 261 was already temporarily halted pending review by the Ninth Circuit last November.  SB 253, or the Climate Corporate Data Accountability Act, requires companies with more than $1 billion in revenue to report scope 1 and scope 2 emissions, with eventual scope 3 disclosures. SB 261, or the Climate-Related Financial Risk Act, requires companies with more than $500 million in revenue to publicly disclose their climate-related financial risks and countermeasures. Both climate disclosure regulations were signed into law in October 2023 by California Gov. Gavin Newsom and have since faced a spate of legal challenges. The ruling came a day before the inaugural emissions disclosure deadline for SB 253. Breaking down the split decision The Nov. 9 decision allowed for the reporting of a company’s direct emissions to proceed under SB 253’s scope 1 and scope 2 disclosure requirements, affirming the lower court’s decision. The court ruled such disclosures “regulate commercial speech,” which the appellate panel defined as factual and objective information on a company’s own operations. However, the appeals court reversed the lower court’s ruling when it came to SB 253’s scope 3 disclosure requirements, which ask companies to report the indirect emissions generated

Read More »

Cisco forecasts $9 billion in AI infrastructure revenue in FY27

Cisco continues to benefit from a “networking supercycle,” reporting record revenue—driven by demand for AI infrastructure, security, and campus refreshes from hyperscalers and enterprise customers—for its fiscal first quarter ended October 24. “In Q1, we delivered record quarterly revenue of $18.5 billion, up 24% year over year, and record non-GAAP earnings per share of $1.40, up 40% year over year. These are our strongest top andbottom-line growth rates in over 15 years,” Cisco CEO Chuck Robbins said on the vendor’s quarterly financial call. “Networking product orders grew 28% in Q1, delivering our ninth consecutive quarter of double-digit growth. In fact, we saw double-digit growth across every product family: service provider routing, compute, enterprise routing, wireless, data center switching and campus switching,” Robbins said. “These broad-based results support our view that we are in the early stages of a multiyear, multibillion-dollar networking supercycle.”

Read More »

How TI’s ADC34DSO75 Catches the Nanosecond Power Glitches Behind AI Data Center Downtime

Today’s advanced AI data center can be viewed as a vast, complex and interdependent digital organism in which a virtually infinite number of components must be tracked all the time, in real time. It’s a monitoring job of staggering proportions. At stake is the operation of highly critical workloads and the maintenance of return on investment for high-cost compute equipment. At risk is sub-optimal performance or, in worst-case scenarios, the nightmare of data center downtime. Sponsored Resources: Why traditional telemetry misses nanosecond power-rail events Telemetry has long been relied upon to monitor data center servers, test and measurement equipment, wireless infrastructure and other systems for real-time information about voltage, current, power, temperature and overall system health. But most systems operate at millisecond-level intervals or slower, which is six orders of magnitude longer than the transient events that now occur at the processor power rails and the tools meant to observe them. The requirements and high stakes of today’s AI data centers and workloads have called for a new generation of advanced, nanosecond-level telemetry gear. Texas Instruments has established a leadership position in this sector. The company has developed the first analog-to-digital converter (ADC) to enable 24/7, low-power, event-based remote monitoring at nanosecond resolution. Let’s unpack this statement and examine why TI’s new monitoring technology is so important for the data center industry. What is the TI ADC34DSO75?    We’re talking specifically about TI’s ADC34DSO75, the industry’s highest-performance, 3 gigasamples per second (GSPS) ADC, delivering thousands of times faster performance than current offerings. Launched at the Open Compute Project Global Summit conference in October, the ADC34DSO75 addresses the shortcomings of existing monitoring solutions that can’t detect the nanosecond-level power-rail glitches that cause costly data center downtime. The ADC34DSO75 combines ultra-high-speed sampling with advanced on-chip digital signal processing in a small form factor,

Read More »

SC26 Preview: AI-HPC Convergence Takes Center Stage in Chicago

As the high-performance computing community prepares to gather in Chicago for SC26, November 15–20, the latest episode of the Data Center Frontier Show offers a timely preview of the technologies, market developments, and infrastructure challenges expected to shape this year’s conference. Hosted by DCF Contributing Editor Doug Black, the conversation brings together Lois McKinnis, Argonne Distinguished Fellow and senior computational scientist at Argonne National Laboratory; Addison Snell, founder and CEO of Intersect 360 Research; and Earl Joseph, CEO of Hyperion Research. The episode also inaugurates Data Center Frontier’s new Frontier Computing channel, expanding DCF’s coverage of the technologies and infrastructure driving the convergence of high-performance computing, AI, QC and next-generation scientific discovery. The central theme is unmistakable. AI is no longer simply an adjunct to traditional supercomputing. It is changing how scientific computing itself is performed. “We’re really moving beyond focusing just on bigger machines, faster machines,” McKinnis observes, pointing to the growing integration of AI into scientific software development, simulations, and research workflows. Joseph underscores the acceleration, citing new Hyperion research finding that 70% of respondents are already using AI for code generation. The panel also examines the U.S. Department of Energy’s Genesis Mission, a sweeping effort to link supercomputers, large-scale AI infrastructure, and cloud resources in support of scientific discovery. Joseph calls Genesis “the first time I’ve ever seen something that’s truly a broad national resource,” while Snell emphasizes that investments in AI infrastructure must be matched by investments in science itself. For data center professionals, the discussion turns to increasingly consequential questions about the economics and sustainability of the AI buildout. Snell warns that the industry’s focus on rapid construction has outpaced serious discussion of infrastructure lifecycles and upgrades. Joseph highlights the mounting difficulty of financing supercomputers costing $100 million or more when new GPU generations arrive

Read More »

Nvidia’s leadership crisis puts enterprise AI’s dependence on one company under scrutiny

“More broadly, it throws into doubt the company’s leadership and security practices, and ignites questions around board-level governance and oversight,” he said. The ‘default choice’ in the market Huang founded Nvidia in 1993, initially targeting the gaming market. But the company pivoted in recent years to meet the intensive compute demands of AI. Its advanced GPUs not only excel at rendering graphics, but can handle an enormous number of simultaneous computations, making them a natural fit for neural networks and LLMs. The company began focusing more on developing chips and architecture for AI tools and data centers, and has continued to ink new deals and establish partnerships with tech companies serving nearly every level of the stack, including Anthropic, Meta, Dell, HPE, Microsoft, AWS, Google Cloud, Accenture, Marvell Technology, Hugging Face and many others.

Read More »

Jensen Huang Resigns as NVIDIA CEO, Sending Shockwaves Through AI Data Center Industry

NVIDIA co-founder Jensen Huang resigned Thursday morning (November 12) as president, CEO and board member of the world’s dominant AI chipmaker, a stunning leadership change that introduces new uncertainty into the global artificial intelligence infrastructure boom. In an official announcement, NVIDIA confirmed that Chief Financial Officer Colette Kress will serve as interim CEO while the board conducts a search for Huang’s permanent successor. The company also appointed longtime director Harvey C. Jones as its first board chairman. The departure ends Huang’s 33-year leadership of the company he co-founded in 1993 and comes amid a rapidly escalating controversy involving his reported personal relationships, potential Chinese intelligence activity and questions about corporate governance. According to Reuters, NVIDIA shares fell more than 10% in premarket trading Thursday as investors weighed the implications for the approximately $4 trillion company and the wider AI technology sector. The company’s announcement did not specify a reason for Huang’s resignation. National Security Investigation Precedes Leadership Change The leadership shakeup follows several days of increasing scrutiny over Huang’s reported relationship with a businesswoman whom U.S. authorities suspect of connections to Chinese intelligence, alongside separate reporting concerning a relationship with a senior NVIDIA executive. On November 8, NVIDIA’s board announced an independent investigation into whether Huang had been targeted by a foreign intelligence operation. At the time, directors stated that they had no reason to believe company secrets or national security had been compromised. The controversy subsequently attracted congressional attention, including calls for Huang’s resignation and questions about semiconductor export controls and the security of advanced AI technology. The espionage allegations remain contested, and no compromise of NVIDIA technology has been publicly established. For the data center industry, however, the immediate question extends beyond the circumstances of Huang’s departure. It concerns the potential disruption of an infrastructure ecosystem increasingly organized

Read More »

Google’s Utah Deal Puts Geothermal Power on the Data Center Map

In Taiwan, the Work Begins With Access to Land Google’s first geothermal procurement agreements in Asia call for an initial 10 MW of new generation on Taiwan’s grid to support local data centers and offices. When Google and developer Baseload Capital announced the agreements in April 2025, a Google energy executive told Taiwan Public Television Service that the companies expected operations to begin in 2029. The original timetable allowed roughly four years between signing a customer and delivering electricity. Baseload’s development work in Hualien County illustrates one reason such projects take time. In November 2025, the company reported that members of the Hongye Tribe had voted to approve a public-land lease for geothermal exploration. Baseload opened a local office, reached an agreement with the community, and secured an exploration permit for a two-year program of drilling in stages. The company has described Hongye separately from Google’s contracted supply. What the Hualien project documents is an early milestone rarely visible in a power purchase announcement: the developer secured community consent before drilling could establish what the field might produce. Kenya Has the Generating Field. The Campus Remains Unresolved. In Kenya’s Rift Valley, the appeal of Olkaria was evident when Microsoft and G42 announced their digital investment initiative in 2024. Kenya already had a fleet of geothermal generating stations at Olkaria. G42 and local partners proposed building a campus there, powered by geothermal electricity and hosting Microsoft Azure services for a new East Africa cloud region. Two difficulties emerged. In May 2026, Semafor reported that President William Ruto had cited insufficient available power as an obstacle to the project. Separately, Reuters, citing Bloomberg, reported disagreements over payment guarantees. Microsoft and G42 had sought a commitment from the Kenyan government to buy a certain amount of computing capacity annually. Kenya’s principal secretary for

Read More »

Microsoft will invest $80B in AI data centers in fiscal 2025

And Microsoft isn’t the only one that is ramping up its investments into AI-enabled data centers. Rival cloud service providers are all investing in either upgrading or opening new data centers to capture a larger chunk of business from developers and users of large language models (LLMs).  In a report published in October 2024, Bloomberg Intelligence estimated that demand for generative AI would push Microsoft, AWS, Google, Oracle, Meta, and Apple would between them devote $200 billion to capex in 2025, up from $110 billion in 2023. Microsoft is one of the biggest spenders, followed closely by Google and AWS, Bloomberg Intelligence said. Its estimate of Microsoft’s capital spending on AI, at $62.4 billion for calendar 2025, is lower than Smith’s claim that the company will invest $80 billion in the fiscal year to June 30, 2025. Both figures, though, are way higher than Microsoft’s 2020 capital expenditure of “just” $17.6 billion. The majority of the increased spending is tied to cloud services and the expansion of AI infrastructure needed to provide compute capacity for OpenAI workloads. Separately, last October Amazon CEO Andy Jassy said his company planned total capex spend of $75 billion in 2024 and even more in 2025, with much of it going to AWS, its cloud computing division.

Read More »

John Deere unveils more autonomous farm machines to address skill labor shortage

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Self-driving tractors might be the path to self-driving cars. John Deere has revealed a new line of autonomous machines and tech across agriculture, construction and commercial landscaping. The Moline, Illinois-based John Deere has been in business for 187 years, yet it’s been a regular as a non-tech company showing off technology at the big tech trade show in Las Vegas and is back at CES 2025 with more autonomous tractors and other vehicles. This is not something we usually cover, but John Deere has a lot of data that is interesting in the big picture of tech. The message from the company is that there aren’t enough skilled farm laborers to do the work that its customers need. It’s been a challenge for most of the last two decades, said Jahmy Hindman, CTO at John Deere, in a briefing. Much of the tech will come this fall and after that. He noted that the average farmer in the U.S. is over 58 and works 12 to 18 hours a day to grow food for us. And he said the American Farm Bureau Federation estimates there are roughly 2.4 million farm jobs that need to be filled annually; and the agricultural work force continues to shrink. (This is my hint to the anti-immigration crowd). John Deere’s autonomous 9RX Tractor. Farmers can oversee it using an app. While each of these industries experiences their own set of challenges, a commonality across all is skilled labor availability. In construction, about 80% percent of contractors struggle to find skilled labor. And in commercial landscaping, 86% of landscaping business owners can’t find labor to fill open positions, he said. “They have to figure out how to do

Read More »

2025 playbook for enterprise AI success, from agents to evals

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More 2025 is poised to be a pivotal year for enterprise AI. The past year has seen rapid innovation, and this year will see the same. This has made it more critical than ever to revisit your AI strategy to stay competitive and create value for your customers. From scaling AI agents to optimizing costs, here are the five critical areas enterprises should prioritize for their AI strategy this year. 1. Agents: the next generation of automation AI agents are no longer theoretical. In 2025, they’re indispensable tools for enterprises looking to streamline operations and enhance customer interactions. Unlike traditional software, agents powered by large language models (LLMs) can make nuanced decisions, navigate complex multi-step tasks, and integrate seamlessly with tools and APIs. At the start of 2024, agents were not ready for prime time, making frustrating mistakes like hallucinating URLs. They started getting better as frontier large language models themselves improved. “Let me put it this way,” said Sam Witteveen, cofounder of Red Dragon, a company that develops agents for companies, and that recently reviewed the 48 agents it built last year. “Interestingly, the ones that we built at the start of the year, a lot of those worked way better at the end of the year just because the models got better.” Witteveen shared this in the video podcast we filmed to discuss these five big trends in detail. Models are getting better and hallucinating less, and they’re also being trained to do agentic tasks. Another feature that the model providers are researching is a way to use the LLM as a judge, and as models get cheaper (something we’ll cover below), companies can use three or more models to

Read More »

OpenAI’s red teaming innovations define new essentials for security leaders in the AI era

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More OpenAI has taken a more aggressive approach to red teaming than its AI competitors, demonstrating its security teams’ advanced capabilities in two areas: multi-step reinforcement and external red teaming. OpenAI recently released two papers that set a new competitive standard for improving the quality, reliability and safety of AI models in these two techniques and more. The first paper, “OpenAI’s Approach to External Red Teaming for AI Models and Systems,” reports that specialized teams outside the company have proven effective in uncovering vulnerabilities that might otherwise have made it into a released model because in-house testing techniques may have missed them. In the second paper, “Diverse and Effective Red Teaming with Auto-Generated Rewards and Multi-Step Reinforcement Learning,” OpenAI introduces an automated framework that relies on iterative reinforcement learning to generate a broad spectrum of novel, wide-ranging attacks. Going all-in on red teaming pays practical, competitive dividends It’s encouraging to see competitive intensity in red teaming growing among AI companies. When Anthropic released its AI red team guidelines in June of last year, it joined AI providers including Google, Microsoft, Nvidia, OpenAI, and even the U.S.’s National Institute of Standards and Technology (NIST), which all had released red teaming frameworks. Investing heavily in red teaming yields tangible benefits for security leaders in any organization. OpenAI’s paper on external red teaming provides a detailed analysis of how the company strives to create specialized external teams that include cybersecurity and subject matter experts. The goal is to see if knowledgeable external teams can defeat models’ security perimeters and find gaps in their security, biases and controls that prompt-based testing couldn’t find. What makes OpenAI’s recent papers noteworthy is how well they define using human-in-the-middle

Read More »