Engagement Farming with AI Bots
Build a fully automated AI blogging pipeline via Gemini CLI, GitHub Actions, Docker, S3, and CloudFront. The bot writes, builds, tests, and deploys one production blog post per day without manual intervention.
John Sobanski
Artificial Intelligence in the Cloud
Latest posts
Crunch Big Data on Your Laptop With Polars Streaming
Polars streaming avoids out-of-memory errors in large cross joins via processing data in chunks. Learn how to run 27M row workloads on a single machine.
Polars Boosted My Algorithm's Speed by 25x
Refactoring an RCE machine learning algorithm from Pandas lambda functions to the Polars expression API reduced execution time from six minutes to fourteen seconds. Polars cross joins, columnar operations, and Apache Arrow drive a 25x speedup.
My Childhood Apple II Computer Programs (1986)
A recovered Apple II floppy disk from 1986 reveals BASIC programs written in fourth grade, including primitive games, low-resolution graphics, and early experiments with loops, strings, and user input. Revisit the code and compare childhood Apple II graphics to modern AI-generated versions.
Love and Food in Hurston's "Their Eyes Were Watching God"
Food references in Zora Neale Hurston’s Their Eyes Were Watching God mirror Janie’s relationships and expose the personalities of her husbands. Cornmeal, sweets, bacon, beans, and breakfast scenes reflect stability, vanity, class, desire, maturity, and love throughout the novel.
Escape the Surveillance Web with Gemini
The Gemini protocol replaces the ad-heavy, JavaScript-driven web with a minimal, privacy-focused alternative built around static content and mandatory TLS. Learn Gemini’s architecture and deploy a Gemini server on Ubuntu with AWS, GMID, TLS, and systemd.
Quick and Easy Capacity Planning with Pandas
Build a lightweight capacity planning model in Python Pandas using flow diagrams, throughput estimates, and GROUP BY operations to estimate CPU requirements and infrastructure cost. Apply Operations Research concepts to size a simple web application without complex simulation tooling.
Why Did Emperor Severus Spare an Engineer?
Roman emperors routinely executed enemies, civilians, political rivals, and even allies without trial, yet Septimius Severus spared the Byzantine engineer Priscus after conquest. Learn why Severus valued Priscus’ technical skill and how the story parallels modern career survival through craftsmanship and reputation.
Combat ChatGPT Use in Academia and Government Acquisition
Detect low-effort AI-generated writing in academic and government acquisition documents via a rubric that considers meme adjectives, passive voice, clichés, empty phrases, adverbs, and unsubstantiated grandiosity.
Mishima's Teens: Adolescent Archetypes of Pain (Part 2 of 2)
Yukio Mishima’s teenage characters embody universal adolescent struggles through identity crises, suicidal idealism, social isolation, and destructive father-son relationships. Explore how these archetype reflect timeless psychological conflict through a read of Mishima’s Sea of Fertility tetralogy.
Mishima's Teens: Adolescent Archetypes of Pain (Part 1 of 2)
Mishima’s teenage protagonists drift through post-Meiji Japan burdened by dread, isolation, pride, and existential emptiness. Honda, Kiyoaki, and Mizoguchi embody adolescent anxiety and detachment across the Sea of Fertility tetralogy.
Install OpenDaylight on Ubuntu 24.04 LTS (Quick and Easy!)
Install OpenDaylight on Ubuntu 24.04 LTS. Perform Java setup, JAVA_HOME configuration, Karaf binary download, controller start, and DLUX enablement. Discover installation traps around Java compatibility and OpenDaylight release packaging.
Wrangle Bearcat BC125AT Input Data with Excel
Transform Radio Reference CSV frequency data into a format compatible with the Uniden Bearcat BC125AT scanner. Explore schema mapping, analog filtering, CTCSS/DCS conversion, frequency normalization, and importing channels into Bearcat programming files.
Use Python Pandas to Invest in Sealed Super Mario Bros. 3
Identify investment grade copies of sealed Super Mario Bros. 3 variants through Python Pandas, Seaborn, and auction sales data. Normalize prices across market cycles and compare box grade, seal grade, release variant, and sale date to rank collectible value.
Beat Artificial Intelligence (AI) with Active, Present Voice
Examine common stylistic patterns in ChatGPT-generated prose, including passive voice, vague abstractions, empty phrases, adverbs, and inflated language. Find clearer technical writing through active voice, Subject-Verb-Object structure, and E-Prime principles.
Data Exploration with Data Viz Cheat Sheet
Witness practical Pandas, Seaborn, and Matplotlib techniques for exploring machine learning datasets using the UCI Abalone database. Includes histograms, KDE plots, boxplots, correlation heatmaps, PCA, regression plots, and multidimensional visualization workflows.
Refactor Matlab to Tidyverse
Refactor a Reduced Coulomb Energy neural network implementation from Matlab into R Tidyverse with pipes, tibbles, functional operations, and vectorized distance calculations. Compares loop-based Matlab patterns with tidy data workflows for training, classification, and visualization.
Roark vs. Raskolnikov: Natural Language Processing Analysis
Compare Howard Roark and Rodion Raskolnikov through sentiment scoring, emotional intensity, and thematic analysis.
Label Speakers in Text with Natural Language Processing
Train a TensorFlow and Keras NLP model to identify and extract Raskolnikov’s dialogue and internal monologues from Crime and Punishment . Perform speaker-level lit analysis via transfer learning, BERT classification, Pandas processing, and sentiment analysis.
Can ChatGPT Write a Decent Literary Term Paper?
Analyze Stephen King’s Carrie through structured prompts, thematic outlines, and historical analogies to evaluate whether ChatGPT can assist with literary term paper development. It explore how GenAI can identify themes, symbolism, and parallels to events.
Test Layer Two OpenDaylight Services In The Cloud
Deploy OpenDaylight and Open vSwitch in a cloud-based layer two environment to steer traffic dynamically with OpenFlow rules. Bypass deep packet inspection for egress traffic while preserving ingress inspection through SDN-based flow manipulation.
An Actionable, Focused, SEO Checklist for Your Website
Practical SEO maintenance for static websites includes sitemap hygiene, canonical validation, redirect cleanup, broken link repair, image optimization, heading structure, JavaScript fixes, and noindex handling for thin content. Learn actionable fixes discovered through Bing Webmaster Tools and Ahrefs audits.
Faux Faulkner Contest: Man vs. Machine
Pit human literary parody against generative AI by rewriting Humpty Dumpty in the style of William Faulkner using modified passages from Absalom, Absalom! . Compare manual collage-style writing against ChatGPT and examine the limits of AI literary imitation.
Written In Blood: Gerard Manley Hopkins & the City (3 of 3)
Gerard Manley Hopkins’ Terrible Sonnet poems I Wake and Feel the Fell of Dark, Not Day, Patience, hard thing!, and My own Heart let me more have pity on portray depression, isolation, and self-imposed spiritual imprisonment brought on by urban life and perceived vocational failure.
Written In Blood: Gerard Manley Hopkins & the City (2 of 3)
Gerard Manley Hopkins’ Terrible Sonnets express spiritual exhaustion, vocational failure, and psychological terror caused by modern urban life. The poems No Worst, There is None and [Carrion Comfort], indicate that Hopkins feared his own inability to fulfill his Jesuit calling vs. doubting God’s existence.
Written In Blood: Gerard Manley Hopkins & the City (1 of 3)
Gerard Manley Hopkins’ Terrible Sonnets reflect despair over urban decay, human degeneracy, and the destruction of nature rather than a crisis of faith in God. Draw from Hopkins’ letters, poems, and biographies, and examine how Victorian city life eroded his faith in humanity and damaged his creative spirit.
Upgrade SageMaker JupyterLab Notebooks to the Latest Pandas
Amazon Web Services (AWS) SageMaker Notebook Instances provide fully managed Jupyter Notebooks, tailored for Data Science and Machine Learning (ML) use cases. These notebooks allow Data Scientists and ML Engineers to explore, operationalize and share data, algorithms and pipelines. Pandas contributes a critical piece to the Data Scientists' toolbox, via …
Jasper Artificial Intelligence (AI) for Marketing Pictures
I use the new Jasper Artificial Intelligence (AI) Art service to create the pictures in this blog post. Jasper AI Art (non-affiliate link), for example, creates the following picture of the World's Largest Turnip . Jasper generates a before and after picture of a fat cat. The AI service also produces …
Turn Your Cellphone into a Secure Shell (SSH) Terminal
Relive the glory days of dumb terminals with your thousand dollar cellphone! Programmers used terminals, decades ago, to log into and execute commands on remote computer systems. The terminal provided a screen and keyboard to computer scientists and phoned home to a multi-million dollar mainframe. Today, we use Secure Shell …
Install OpenDaylight on Ubuntu 22.04 LTS (Any Version)
Internet Service Providers (ISP), Cloud Service Providers (CSP), Data Center Engineers and Academics use the OpenDaylight (ODL) platform to tailor and automate computer networks. Professionals and Academics use ODL to execute several use-cases: Network Service Delivery Automation Use an Application Programming Interface (API) to provision network connections and virtual private …
LTE: Hierarchical to Flat IP Mobile Architectures (Part 2)
This month's blog post continues to discuss the evolution from hierarchical to Flat Internet Ptotocol (IP) Architectures for mobile networks. Be sure to read Part One if you have not already. This post briefly discusses the components of mobile data technology evolution and lists the strengths of a flat Internet …
LTE: Hierarchical to Flat IP Mobile Architectures (Part 1)
I wrote this white paper a while back. This paper discusses the evolution from hierarchical to Flat IP Architectures for mobile networks. This information may be of interest to mobile network architects, especially in light of the emergence of Software Defined Radio Access Networks (SD-RAN). Introduction Consumption of IP data …
Connect your 8BitDo M30 Bluetooth Controller to RetroPie
I purchased the new 8BitDo M30 Bluetooth controller for my System76 (Ubuntu) Laptop. I connected the M30 to RetroPie and then noticed issues with button layout for certain Six (6) Button Sega Genesis games. In this blog post I will walk us through how to connect the new controller to …
Jasper Artificial Intelligence (AI) Writes Your Blog Posts
Machine Learning (ML) Engineers at Jasper Artificial Intelligence (AI) (non-affiliate link) trained a Natural Language Processing (NLP) model on 10% of the written content on the Internet to create a service that writes text on command. A user feeds the model a sentence or two requesting a topic and the …
Forward Error Correction Visualized - Convolutional Codes
Over the air communications, such as text messages, satellite radio, walkie talkies and WiFi need to deal with the unpredictable effects of noise in the channel. Coding theory attempts to make noisy channels more reliable by bringing the noise down to a theoretical lower bound of zero. Channel coding, as …
Forward Error Correction (FEC) Visualized - Block Codes
Engineers face the reality of imperfect and noisy channels when designing digital communication systems (DCS). The engineers must account for signal degradation to ensure that a receiver receives the message that the transmitter intended. Consider the classic kids’ game telephone, which provides a familiar example of a noisy channel. The …
A New Exemplar Machine Learning Algorithm (Part 2: Optimize)
In part one of this two-part series, I developed a Reduced Columb Energy (RCE) classifier in Python. RCE calculates hit footprints around training data and uses the footprints to classify test data. RCE draws a circle around each labeled training observation, with a radius (lambda) that stops at the closest …
Host a 100% Decentralized Website on the Ethereum Platform
The free and open Ethereum platform delivers the following benefits to developers and end-users: Impervious to censorship Native payment system Universal, plug and play Database Global, anonymous login 100% Uptime This blog post demonstrates how to use InterPlanetary File System (IPFS) with Ethereum Name Service (ENS) to host a static …
Darkblock Levels Up Non Fungible Tokens (NFT)
Darkblock provides a decentralized application that lets creators sell (and manage) digital content directly to consumers without the need for any intermediate, proprietary or bureaucratic services. Their splash page touts content access controls for the decentralized creator economy . From a tactical perspective, Darkblock allows the creators of Non Fungible Tokens …
Get Brave Creator Verified and Earn Creator Rewards
Brave Software developed a privacy-focused browser, that (unlike competitors) blocks all online advertisements and web trackers by default. Brave Rewards pays out content providers crypto based on views. In order to get paid, you must get Brave Creator Verified . A blue check indicates if Brave granted a given website Brave …
A New Exemplar Machine Learning Algorithm (Part 1: Develop)
In Pattern Classification Using Neural Networks (IEEE Communications Magazine, Nov. 1989) Richard P. Lippman provides the following definition of Exemplar neural net classifiers: [Exemplar classifiers] perform classification based on the identity of the training examples, or exemplars, that are nearest to the input. Exemplar nodes compute the weighted Euclidean distance …
Adaptive Forward Error Correction (AFEC) for the Ka band
Satellite system engineers consider the narrow beams (which lend themselves to frequency reuse) and high data rates of Ka band channels attractive for communication systems [ Khan 2]. As a result, the satellite community witnessed a great increase in the number of Ka band Satellites over the past two decades. During …
Juice Your In-Situ Machine Learning with BigQuery AutoML
Data Scientists need skill and experience to create useful Machine Learning (ML) models. ML activities include tool selection, training logistic decisions (move data to training vs. train in-situ), data acquisition, data cleaning, data quality checks, feature engineering, algorithm selection and hyperparameter tuning. Algorithm selection and hyperparameter tuning drive tedious manual …
Did Thoreau inspire the Unabomber? We use AI to Find Out!
Good Vs. Evil - Two Opposing paths Taken by a Similar Genius This blog post provides a comparison between Henry David Thoreau's Walden and Ted Kaczynski's Unabomber Manifesto. To compare these two works, I use both a modern Natural Language Processing (NLP) Artificial Intelligence (AI) tool and traditional literary analysis. The …
The Non-fungible Token (NFT) Value Prop
NFTs provide an immutable record of ownership for digital assets, verified by the global Ethereum Blockchain. The Blockchain tracks ownership of NFTs until the end of human civilization. 2021 ushered in speculative mania in every conceivable investment and collecting category. Real Estate, Stocks, Precious Metals, Comic Books, Pokemon Cards, Graded …
Create Your First Non Fungible Token (NFT) w/ Rarible Part 2
Rarible enables artists to mint, sell and trade one-of-a-kind digital collectibles in the form of Non Fungible Tokens (NFT). The Ethereum blockchain records the provenance of each NFT in a secure, distributed ledger that will last until the end of human civilization. You don't need to be a Blockchain wiz …
Create Your First Non Fungible Token (NFT) w/ Rarible Part 1
Non Fungible Tokens (NFT) allow collectors to own unique, scarce digital assets. NFT protocols record ownership and provenance on the Ethereum Blockchain. Recent offerings from Jack Dorsey, Logan Paul and Lindsey Lohan helped fuel the current mania for digital collectibles. The NFT landscape in early 2021 attracted manic investors, eager …
Don't Move Your Data! In-Situ Machine Learning via BigQuery
I started my AI/ML journey in 2011 with a laptop model , a term which indicates a measure of size. Laptop models , by definition, do not exceed the compute, memory and storage resources of a single piece of hardware. The laptop model approach works well for small data sets, and …
Fast and Easy ML Optimization with GCP AutoML Tables (Beta)
Model optimization on traditional Artificial Intelligence and Machine Learning (AI/ML) platforms requires considerable Data Architect expertise and judgement. These ML platforms require the Architect to choose from dozens of available training algorithms. The platforms also provide a host of hyper-parameter knobs and switches for the Architect to tune. The …
Fast & Easy Regression w/ Keras and TensorFlow 2.3 (Part 2)
In this demonstration we continue to use Keras and TensorFlow 2.3 to explore data, normalize data, and build both a linear model and Deep Neural Network (DNN) to solve a regression problem. Today we use Principal Component Analysis (PCA) to address over-fitting via dimensionality reduction NOTE: TensorFlow Core 2 …