do not look into laser with remaining eye
- Posts
- 910 from 13 Mar 2011 to 16 Oct 2026
- Posting
- weekly (24 in the last year)
- Kind
- personal blog
- Feed
- https://graphthinking.blogspot.com/feeds/posts/default
2026 22
- interactions with LLMs cannot feel awkward (currently)
- lessons from implementing research workflows with small local LLMs
- imbalance for cybersecurity and AI
- design of a multi-agent team
- refinements to considerations for multi-agent teams
- considerations for multi-agent teams
- desirable conversation features insights
- software project folder structure for isolating agent scope
- categories of AI for helping onboard folks who are new to AI
- multi-agent teams of LLMs
- lessons from enacting agentic workflows for software development
- interview questions
- second-hand sources of information
- Personalized feeds are harmful
- from LLM chat to agent to guided workflow
- example scaffolds for AI prompt orchestration to navigate complicated multi-step workflows
- synchronization of buckets filling and emptying
- cargo cult bureaucracy
- historical comparisons for how AI is revolutionary to how businesses operate
- snapshot of my inability to determine veracity for topics I'm not familiar with
- can't things just be better?
- probabilities (don't have to) multiply for agentic systems
2025 19
- the persistent need for general knowledge and subject-matter expertise with large language models (LLMs)
- we have trained machines to dream
- Prompt engineering advice and how to review output from LLMs for text-based interactions
- why we have large complicated bureaucratic systems
- Bureaucracy as a compounded principal-agent problem
- Bureaucratic livelock
- how to make money as an LLM service provider
- visualizing family tree
- plaintext genealogy book to neo4j cypher query using Gemini 2.5 LLM
- Document the motivation for creating a Policy, Standard, or Procedure (PSP)
- Gradations of the Inverse Conway Maneuver
- Unspoken Currency of Office Politics: Synergy and Punishment Between Coworkers
- scaling my management tactics
- buying a house for the first time
- documenting ineffectiveness of an employee in preparation for firing them
- Jira best practices for task tracking
- movie and TV show tropes in scripts
- monte carlo addition in Excel as per Gemini 2.0 LLM
- knowledge hidden in legacy code
2024 41
- uses for LLMs
- Makefile as wrapper versus abstraction layer
- Tracking tasks tactically versus strategically
- distinguishing passive from active, busy from productive for knowledge work
- concurrent processes in bash
- search text corpus using RAG with an offline LLM
- Founder mode and Dunbar's number
- characteristics of kindness
- observations too small for their own blog post
- task coordination: benefits, costs, and risks
- Reporter, Interpreter, Manager and Educator
- emotional complexity and a segmented audience
- splitting versus aggregating talent and resources
- enabling cohesion and productivity
- doing delegation right
- commander's intent
- a multitude of prioritization tactics
- Site Reliability Engineering
- analog for management or leadership
- sources of innovation and invention
- my conception of leadership
- Being an effective middle manager
- text editors for Mac in 2024
- maturing as an employee
- emotional response to logistics of email
- relation between goals, vision, strategy, tactics, risk, costs
- Envisioning effective Knowledge workers in an office
- why re-org -- motives for changing the structure of an organization
- leveraging the middle majority
- tools managers have for interactions with people they manage
- an evolutionary story of bureaucracy
- notes from interview with Alexandria Ocasio-Cortez
- how bureaucracy intersects with various economic and governance paradigms
- revising my definition of bureaucracy
- Scaling plateaus of distributed decision making
- bureaucracy: fiefdoms
- coordination unreliability in a bureaucratic organization
- why managing a team of software developers is challenging
- free podcast apps for Android in 2024
- what is risk -- analyzing an overloaded term
- Prioritization of tasks from first principles
2023 41
- next steps for LLMs -- working in flocks and leveraging resources
- augmenting a containerized large language model (LLM) with text files for context
- Mistral 7B LLM (5.6GB) with llama.cpp in a Docker image
- small offline large language model -- TinyChatEngine from MIT's Han lab
- efficiencies enabled by LLM will be less than expected
- book ideas that are currently just blog posts
- measuring the dysfunction of an organization
- The axes of a matrixed organization should not be arbitrary categories
- Spectrum of mismanagement patterns
- relevance of coworkers in an organization
- How to screw up design of a matrixed organization
- cost centers are more political and members feel less empowered compared to profit centers
- more than just training to gain skills
- busyness versus productivity
- patterns and anti-patterns of effective decision makers and communicators
- Is what I am doing relevant for my organization?
- why bureaucracy feels bad
- matrixed organizations: best practices and risks for team members at the bottom of the hierarchy
- Expecting equitable contributions from team members
- Skills relevant to leaders of organizations
- What do successful team members do?
- Keywords: fire, firing, terminate, termination
- Boxologist
- Responsibilities and Activities of a Manager
- pdb versus tracing in Python
- valgrind instruction count for Python on bare metal and inside Apptainer
- vTune doesn't run on DigitalOcean droplet
- what files does this program touch? Using strace to track files bare metal and inside Apptainer
- considerations I didn't predict before becoming a manager
- misalignment of tasks
- "I want to continue to learn" as an aspirational statement by newly graduating job applicants
- Do my actions and investments align with my priorities?
- how to spend my attention on what I want and what others want from me
- POSIX+certificates versus attribute-based access control (ABAC)
- Running Python in a Singularity container using Podman
- Python3 using podman container on CentOS 8
- Prioritization of tasks
- Differentiating unexpected events using local information
- Atlassian Jira jargon and definitions
- Jira projects: just one project versus multiple projects
- More Questions to ask when you encounter a Problem or have an Idea -- Beyond Heilmeier's Catechism
2022 23
- Podcasts worth listening to again
- collaborative war-gaming for planning research activities
- coming up to speed: how to learn a new domain
- deceptively harmful attitudes held by individual contributors on software teams
- webex tip for tones
- matchmaking for mentoring
- dumb idea: performance improvement using random fuzzing
- questions to ask in a one-on-one meeting
- Enumerating the specific challenges that result in every big thing being difficult
- failing fast when writing software: how to detect failure?
- gitignore -- to limit my own mistakes
- periodically monitor memory, CPU, disk writes, network on linux using SAR (system activity report)
- encrypted partition on CentOS 7
- SSH key authentication for multiple github accounts
- Automating Tasks for Processes that Involve Intangibles
- website idea: explain C and C++ (like explainshell.com)
- bureaucratic processes typically involve a person new to the process
- escheat websites
- periodic extrinsic rewards for team members -- traveling trophies, employee of the month
- managing a Python code base
- notes from a professional development course on public speaking
- Rove R2-4K dashcam diagnosis of wifi
- growing a team
2021 97
- Control and trust are harmful models for interaction
- what is not working at a societal level: lack of feedback loops for assholes
- hierarchical organization trilemma: individual autonomy, peer collaboration, and accountability-to-supervisor
- impediments to creating a sense of ownership for members of a team
- questions to ask distant relatives or close friends of the family that you have not interacted with in a long time
- brainstorming with other people can challenging; how a facilitator can respond to situations
- sick 2021
- determining the size of an electrical power outage using locally available information
- comments on arxiv papers: multiple options
- management behaviors that don't scale well
- block uploads of certain files to a shared bare git repo using update git-hook
- strategy development in relation to vision and planning
- this plane can fly: a dependency graph of claims
- structuring email content for professional communication
- documenting a software interface of an SDK
- why I don't like being in a management role
- why is everything so hard in a large organization?
- notes from a class on being politically savvy (in terms of organizational politics)
- blob of code as default design; decomposition with well-defined boundaries is unnatural
- headless execution of Jupyter notebooks on the command line
- notes from a short course on leading four generations
- all code committed by a person to a git repo
- notes from a half-day course on coaching and feedback for managers
- Human failure modes and plateaus; behavior that distinguishes dumb from smart
- brew in Docker
- ways to re-engage in a negative relationship
- Thought terminating concepts in planning for a team
- lack of automatic parallelization for scientific computing software used on High Performance Computers
- bureaucracy book outline
- torrents in Docker using Alpine and transmission
- visualizing the Top500 supercomputing lists
- Warren Bennis on leadership -- notes from articles
- screening for intellectual empathy in job applicants
- frustrated with my team
- machine learning from scratch - website resources
- Using bridges2 HPC via SSH from the XSEDE portal for MPI jobs
- parameter exploration in bash with subshells and containerization
- what domain should my second PhD be in?
- Patterns, anti-patterns in bureaucracy of large organizations
- three measures of bureaucratic maturity
- purposes of promotion
- finding the most recurring topics on Hacker News
- questions for non-fiction book club
- my preferences on communication at work
- definition of done for reports and software with iterative milestones
- technical responses to insecurity when developing software
- Periodic rotation across projects to diffuse ownership within the team
- not my job: task scope and collaboration
- git hook for pre-commit -- to limit my own mistakes
- software that is easy to refactor
- span of documentation; documentation should include assumptions and limitation
- historical progression of organization goal setting: comparison of MBO, S.M.A.R.T., KPI, OKR
- the agenda for a one-on-one meeting between supervisor and employee should evolve with the employee
- software development observations inspired by "Turn the Ship Around!" by Marquet
- how to work towards promotion within your role
- a spectrum of bug reports and instructions
- Reading list for Professional Development
- team members as volunteers rather than employees
- Management strategy: how to allocate my attention time
- Unreproducible work causes harm when resources are zero-sum
- Laffer curve and minimum viable bureaucratic processes
- notes on the FIPS "Guideline for Software Documentation Management"
- questions to ask the interviewer when applying for a position
- identifying how granular a modular design should be
- What leverage does a team lead have in interactions with team members?
- repeatable patterns implementing metacognition through specific activities
- screening for metacognition in job applicants
- evolving into a research mentality
- offloading the task of leadership to team members
- responsibility for societal-scale process debt
- Questions each team member should be able to answer
- dynamics of a cohort versus a team, and when to use which
- a tension between working at a discount rate versus promoted to the level of incompetence
- read each email to determine the purpose
- service-providing businesses, plug-in frameworks and the Unix philosophy
- given an unknown, create a plan
- reduced version of Heilmeier questions for readback
- what I dislike about being a manager
- employee handbook is about expectations
- attributes relevant to being selected as a leader
- Greek key motifs - patterns in the maximal space-filling curve on a two dimensional (2D) Cartesian grid
- path notations for the maximal self-avoiding walk on a two dimensional (2D) Cartesian grid
- notes from Hamming's lectures
- Greek key = maximal self-avoiding walk = a superset of the domino tile covering challenge
- what to do about dirty addresses in databases
- team lead serves to resolve blockers for individual members and the team
- organizations value things more than people's time
- How to have an efficient bureaucracy
- active shaping of culture is essential within a team
- Personalized micro surveys using outlook mail merge
- a routine for knowledge discovery within a research group
- what drives organizational structure and what should
- paradoxes for a project manager to internalize
- map of graph data structures and distinguishing features
- data structure for composition in hypernode graphs
- category theory: notes on operads
- isolating the difference between a working SemanticMediaWiki installation and my broken process for extension installation on MediaWiki
2020 69
- UML diagrams for classes in Python scripts using Pyreverse
- off-line installation of Semantic MediaWiki by creating an individual file release
- Composer-based installation of Semantic MediaWiki extension with MediaWiki running in Docker using docker-compose
- MediaWiki with MariaDB in Docker using docker-compose
- unable to stay in a Days Inn hotel room that I paid for because I did not show up on the first night of the reservation
- save a running Docker container as an image, then save image to file
- mediawiki in Docker, with (ephemeral) Semantic MediaWiki using Composer
- why is cooking an uncommon skill
- strategy for keeping up with notes at work
- impact of self-identifying as not a bureaucrat
- Meals that I've eaten at work -- frozen, refrigerated, and fresh
- my weekend mealprep routine
- Who is free?
- Singularity in Ubuntu (since I can't build using Singularity v0.1 on Mac)
- building a mental model of code: how to decrease barriers
- mentat training sources
- democracy is premised on intellectual conflict being desirable
- tactics for warmness -- being warm as a set of behaviors
- systemic equilibrium of specialization as a function of scale
- wealth flows upward in a hierarchical organization
- mistake log
- replace "don't do X" in feedback with constructive analysis
- why migrating from current to new technology is hard
- survey of tactics for addressing document length and complexity
- generators and coroutines and asyncio in Python
- Identifying and Eliminating Unintentional Ambiguity in Written Communication
- reproducible CPU and memory scaling tests for Python using Docker, bash, and Make
- notes from initial effort to learn javascript
- what skill deficiencies are restraining my progress
- assume noble intent applies only to initial interactions
- how to have a useful conversation
- best practices for tracking experiments implemented in software
- personal, emotional vulnerability and professional, processes vulnerability
- visualization software comparison
- the model and assumptions are rarely presented in software documentation
- initial experience of separate work and home life
- scope creep is experienced differently for customers and creators
- abstraction layers, models, assumptions
- choice of data structure used in communication enables or inhibits progress
- What to ask instead of "Does anyone want to volunteer for this task?" to get engagement
- risk associated with simplifying assumptions in storytelling
- aspects of software to optimize
- hearing a lie causes harm to the listener
- create a JSON file based on a JSON schema using a web interface
- what skillset is relevant in the age of misinformation and misdirection
- framing causes misdirection of attention
- what is the relevance of a Pareto frontier
- github regex search
- searching through github issues using the API
- checklist for commuting to work
- scale versus complexity
- invisible bureaucracy
- complexity is measured by number of abstraction layers
- Causes of Software Bugs
- toilet paper purchase behavior as a measure for emotional duress associated with the pandemic
- prediction for hospital capacity
- behavior changes associated with social problems
- sequence of events for the COVID-19 pandemic
- software archeology for legacy code
- a prediction of the cultural response to the pandemic
- My experience during the pandemic
- identifying traps used by narcissists
- design for turnover rather than rely on heros
- there is no "idle" status for paid employees
- search interrupted - Google foobar test for level 2 - gears
- search interrupted - Google Foobar test for level 1 re-id
- summary of HackerNews recommender
- Hierarchy of justification
- the consequences of stupidity
2019 74
- Goals
- a process I want to avoid being involved
- what is enlightenment?
- formal vs informal processes
- I love you. So what?
- Consequence of normalizing video games
- what is the purpose of this meeting?
- Notes from Key Elements of Machine Learning
- formal education for domain exposure versus learning to learn
- software is held to a higher standard
- new jobs of the future
- tesseract OCR in Ubuntu Docker container
- imagemagick convert PDF to PNG in Ubuntu Docker
- summary of points from Herb Cohn's book "You Can Negotiate Anything"
- what I want in a job
- two misleading simplifications when reasoning about large-scale systems that involve humans
- brainstorming career options in medical research
- from informal conversations to social interaction
- first principles analysis of creating organizations from scratch
- cost of creating and supporting automation
- bureaucracy is a social process executed by humans
- vulnerability of organizations in describing lessons learned
- how to rephrase negative observations into constructive positive statements
- winning the game
- not too loose, not too tight: determining scope in an organization
- altering feedback loops to change society
- human organizations and emergence
- quantification of human interaction: attention-time
- being the subject matter expert can feel isolating
- attributes of information
- cgroups on Ubuntu
- comfort, risk, happiness, sustainability
- understanding fascism
- benefit and harm of corporations in a market
- enumerating the benefits I derive from participating in society
- standard issues that an organization must address
- improving the skill of prioritization
- prioritization of reading articles when there is insufficient time
- Filter questions for finding interesting and compatible people
- Responding to what a person says -- three ways
- categories of people associated with machine learning
- micromanagement
- enumerating options for changing society
- labeling comes with emotional associations
- categories of conversation
- miscommunication versus inability to communicate
- addressing violence by determining underlying causes
- three games for improving communication
- fail fast versus quick and dirty
- cooperation and competition
- definition of progress
- morals in data science inform my career: I don't want to support dark patterns
- how I respond to differences
- where to volunteer
- environmental debt
- the badness each person hides
- identifying empty talk
- Role of Media in Capitalist Society
- Role of Feedback loops in Process Design
- dealing with rules
- notes from a salon on privilege
- Exploring Erlang
- anti-superficial rather than antisocial
- seek counter evidence to what you believe
- emotional satisfaction tied to perceived progress
- sequence of failure due to electrical power outage
- attention is the new commodity
- categories of people associated with machine learning
- vision for a better task management system
- political decisions versus science-based decisions
- Threshold for Forgiveness; not forgotten
- Time-to-first-solution versus time-to-recurring-solution
- Why science is important to making progress
- anonymous student peer review
2018 83
- status of analyzing recorded verbal communication; milestones for cleaning audio file
- creating an app for Android
- mediawiki docker image with SQLite
- a GoogleWhack
- structured brainstorming versus free association brainstorming
- lack of documented communication causes organizational friction
- windows disk images and virtual machines
- notes my limited exposure to genetics
- connect jupyter running in Docker to Neo4j running in Docker
- how I process human communication
- ubuntu docker with gprof and gcov and valgrind
- managing a cluster that has a single user
- skills for human interaction in meetings, leadership, and bureaucracy
- consequences of trying to figure out how humans work
- introspection about how I work
- in-class activities in a data science class
- rubric for grading writing assignments and coding assignments
- Innovation causes disruption
- relevance of scrum for dysfunctional teams.
- idea: tool for identifying conflicting definitions
- why incur technical debt, and the consequences of incurring tech debt
- how to communicate poorly
- standard homework responses when grading assignments
- peer review of essays and reports during class
- correlation of funding and votes on bills
- lessons learned from assigning my first essay assignment
- dealing with Warnock's dilemma in the classroom
- management thrashing
- getting Python 3 kernel in Jupyter on Mac OSX
- interactive class activities to encourage engagement
- variables of interaction in the classroom
- tips for fostering an engaging class
- data science reading list
- sequential approval chains are inefficient and harmful
- categorizing social conventions
- distinguishing conversational story telling from contemplative introspection
- how to decrease bureaucracy
- confusion leads to confusion
- first phone interview for adjunct faculty position
- fear of failure and fear of iteration
- modeling human interactions
- emotional improvement and games
- geographic data for the USA
- two hurdles for process automation
- Evolutionary sequence for data management
- family tree in Neo4j for genealogy
- the difference between problems at school and problems at work
- operating with bureaucracy: by exception and by routine
- speech to text milestones
- passion
- Thoughts and feelings filter information
- increasing the precision of how I understand language
- imprecise language
- data driven decisions versus data-inspired decision
- examples of societal-scale anarchy versus structure
- interrupt-based competition for verbal interjection
- extracting audio from mp4 on Mac OS X
- how to dominate in the infinite game
- OCaml in a docker image
- my evolving view on the role of my reputation within a large organization
- idea for audio machine learning app: categorization of baby crying for new parents
- observations on organizational politics
- install bash from source on Mac OS
- enumerating knowledge as triples
- incompetence versus fraud in government contracting for science
- aspects that determine whether I sleep well
- literature review for thinking about complex bureaucracies
- why I work
- cost/benefit evaluation of randomly selection an option versus investigating choices
- "don't disagree with me after I decide" reduces sources of innovation
- granularity of documentation and results for experiments
- useful predictions depend on useful data
- inconsistency and documentation
- Neo4j and Jupyter
- idea: coloring book of cuddling, hugs, and communication wisdom nuggets
- Elasticsearch in docker containers on a cluster using IP over Infiniband
- conversation openers for finding people I want to spend time with
- Poor communication - a case study
- analysis of "A follows B" data using bash
- "Who benefits?" applies to individuals, not groups
- indicators of effective communication and poor communication
- where (and why) to give my money
- Entertainment
2017 131
- culture
- I hold written content to a different standard
- collected wisdom
- the value I find in interacting with other people
- fast filtering of new social interactions to find interesting people
- laws are subjective and dynamic
- balancing ad hoc response and reliable service
- research on poverty reduction
- review of types of donations
- Criteria for giving money sustainably to reduce poverty
- research organizations for reduction of systemic poverty
- how I want to donate money
- visualizing audio data -- amplitude (volume) and spectrograms
- growth via risk based on what I value
- ceremony and ritual displace original, personal thinking
- crosstalk in social interactions
- a list of colloquialisms
- growing mushrooms indoors
- upcoming medical conferences
- Getting started with TensorFlow
- why platitudes are used and why they should not be
- multiple definitions of vulnerability
- moving big data and the cloud: my experiences
- a list of platitudes
- update all my outdated Python packages
- How to respond to being wronged
- seeking optimization and seeking emotional congruence
- My current model of human behavior
- What was my motivation to travel?
- the difficulty of improving
- decision tree for change in your environment
- flipping the approval mentatlity
- maximum market size and maximum bureaucracy size doesn't exist
- complex tasks necessitate bureaucracies or markets
- complex tasks addressed by market versus bureaucracy
- bureaucratic debt and what to do about it
- market friction and bureaucratic inefficiency are equivalent
- a simple illustration of bureaucracy
- enumerating potential employers
- history of machine learning
- engineering my social interactions
- how do you respond to adversity?
- cProfile in Python for finding application hotspots
- what can you risk at work?
- how to be vulnerable at work
- why is my computer slow? A decision tree
- easy prediction: Augmented Reality will dominate over Virtual Reality
- motivation and measurable features
- heart rate, heart rate variability, and beat-to-beat heart rate
- using linear algebra to place domino tiles on a grid
- linear sequences of domino tiles
- simulating a bureaucracy
- inventory of ideas for tiling a floor with dominos
- scaling of transition table breadth first search
- multiple representations for tiling a 2D grid
- feelings and action
- single pane of glass
- a histogram of characters in a text corpus
- Simulation environment: OCCAM + SST
- Technology and progress
- Emotional relevance of order of events
- generalization versus the big picture
- if I have everything I've been aiming for, then why am I not content?
- analyzing text documents for knowledge extraction: detecting delimiters
- scaling up a random domino tiling
- a basis for analyzing the randomness of a tile floor with dominos
- software programs of increasing complexity
- questions to ask mentors
- features of people I want to socialize with
- sum 1 to N using a for loop in multiple languages
- Enabling asymmetric advantage
- Measuring the growth of a bureaucracy
- Examples of quantitative cost benefit evaluation
- food and emotions
- stuff needed for 200 people at a wedding
- what can be done about email
- importance, impermanence, imperfection
- Ubuntu packages and settings I used in 2012
- population sizes needed to support innovation
- What will drive humans to leave Earth?
- Elements of an enjoyable meeting
- comparing methods for topic modeling
- presence with ( attention * time * alertness ) in interactions
- Options in the decision making process
- reproducible data science
- a histogram of bash history
- Facebook development
- Building a neural network from scratch in Python
- counting x86 instructions in a binary
- tropes in games
- patterns: aperiodic, periodic, random, chaotic
- entropy examples
- growth of bureaucracy
- progress evaluation at work
- static versus dynamic processes
- progress in spite of better ways
- what slows down bureaucracy?
- intentionally shifting my social circle
- 3 options, choose 2
- the meaning of my life
- NLTK sentences and Named Entity Recognition; SyntaxNet Docker image in Ubuntu
- community, teams, and hierarchy
- data driven decisions are insufficient
- logging into an AWS EC2 instance over SSH
- install sklearn on Ubuntu
- relation between task complexity and when to use a different tool
- Breaking down bureaucracy: process, roles, people, machines
- financial motivation in bureaucracy
- abstraction layers: ease of use versus risk
- schedules slip
- getting deeper faster in social interactions
- phrases which serve as thinking stoppers
- signal interrupts to my interactions
- SyntaxNet Docker on Ubuntu host
- Docker on AWS EC2 Ubuntu
- role of creativity and productiveness in work
- communication in a large hierarchical organization
- inventory of emotions, feelings, physiological sensations
- data, information, knowledge: a picture
- practical transformation of information to knowledge
- social meditation in real life without specifying expectations
- Measurable impact of poor user interface design
- Three distinct ways to make intentional progress
- (more education, more information) are not sufficiently specific to be actionable
- observations during meditation: suffering
- home inspection checklist prior to buying a home
- set theory assumes classical mechanics and is time invarient
- sailing club logistics - finding a crew
- financial limits and recommendations
- buying a house: tips and advice
- Observations during meditation: emotions, sensations, signals
2016 153
- inventory of meditation practices
- linear storytelling is a walk along a knowledge graph
- tile a floor with dominos
- why and why not to have kids
- options for dealing with the worry about leaving my stove on
- Observations during meditation: Avoiding surprise and frustration, Loving Kindness, Ego
- not clear how to learn lessons in large organizations
- life lessons I learned from experience
- repost from Google plus on Service Oriented Architecture
- Silos versus Service Oriented Architecture
- Hacker News recommender, part 3 -- a toy array of data
- causes of stress at work
- image processing combinatorics and information theory
- feedback on an attempt to get started with neural networks
- Feelings log
- image classification using MNIST images
- Words histogram on the command line from the text of a webpage
- docker on Mac OS X
- energy use for gas versus electric vehicles
- notes on good conversation
- Convert text in image to text using OCR
- to increase your financial security, implement a credit freeze
- justified paranoia should lead to action
- work: what I want, what I don't want, and my history
- dissecting morals
- reflecting on a mistake leads to insight on evolution
- Awkward stories as a route to establishing deeper interactions
- learning through experience: mutable and immutable data types in Python
- Microsoft's cloud software offerings
- charge models
- why are processes less efficient than they could be?
- game theory and economics as routes to understanding politics at work
- to avoid politics, I'm going to use measurement
- being genuine and feeling compassion
- Scraping Hacker News articles, part 2: comparing posted to what I read
- Scraping Hacker News articles, part 1: plan and getting data
- getting Chrome browser history
- how to interact with someone who processes actions differently
- scales of interaction in large organizations
- day 10 of 10 | Introduction to C++: standard template library (STL)
- measuring factorial growth of permutations on a list in C++
- C++ Jeopardy
- Introduction to C | Day 8 of 10 | overloading "<<"; command line arguments
- evolving faster software applications
- Introduction to C++ | Day 6 of 10: classes; overloading operators; uninitialized pointers; exit
- Introduction to C++ | day 5 of 10: constructors
- if everyone is doing right, then why are there suboptimal outcomes?
- Introduction to C++ | Day 4 of 10: scope resolution operator; reference parameters
- Introduction to C++ | Day 3 of 10: source to binary; call header once; structs
- Introduction to C++ | Day 3 of 10: fizzbuzz, profiling using valgrind
- Introduction to C++ | Day 2 of 10: polymorphism, header files
- Introduction to C++ | Day 1 of 10: C vs C++ hello world, name space, make files
- three process types in large complex organizations
- service model and charge back in a non-profit monopoly
- proposal for a closed betting pool on process outcomes
- install TensorFlow on Ubuntu and no GPU
- machine learning from scratch: computational kernels
- download historical local weather data
- Dominance of Excel in engineering, data analysis, data science, and many other fields
- using FlameGraph on Ubuntu for simple tasks
- CPU instruction count in Ubuntu
- CPU instruction count on Mac OS X. Spoiler: I didn't get it to work :(
- when to use which machine learning algorithm
- machine learning from scratch: predicting mpg for automobiles
- machine learning from scratch: preview numerical data using visualization
- machine learning from scratch: separating data into test and train
- machine learning from scratch: data cleanup process
- there's more than one way to calculate the mean; when to use which mean
- graph analysis of website
- what level of granularity is needed for monitoring of data center resource consumption
- ASHRAE 2016 St Louis meeting notes
- machine learning from scratch
- patterns in code
- decreasing risk through investment
- Why human interactions are "messy"
- the danger of optimizing for efficiency: research and failure
- how to write everything down
- innovation: in the open versus behind the curtain
- Introduction to C: course notes | day 10 of 10: useful libraries
- notes on advanced linux shell
- Introduction to C: course notes | day 9 of 10: read from file
- Introduction to C: course notes | structures, typedefs, bit fields
- Introduction to C | day 8 of 10: pointer to character pointers
- Relation between data scientists, production engineers, and customers
- Inventory of Software Tools for Data Science
- chaos monkey as forcing function for availability
- Introduction to C: course notes | day 6-7 of 10: pointers
- impact of requests via back channels
- Introduction to C: course notes | day 6 of 10: libraries and header files
- facts and emotions
- Introduction to C: course notes | day 5 of 10
- differentiating "good leadership" from "good person"
- Introduction to C: course notes | day 4: assignment
- Introduction to C: course notes | day 4: file I/O
- Introduction to C: course notes | day 4: #define
- Introduction to C: course notes | day 3: scanf and loops
- Introduction to C: course notes | day 2: data types and sizes
- top-down and bottom-up approaches to change in hierarchical organizations
- Introduction to C: course notes | day 1: compiler essentials, data types
- not a leader and not a manager
- principles of privacy
- scaling of organization is constrained by processes which involve humans
- computational cost of transformation of data
- how innovation can fail
- claim: innovation is either disruptive or iterative
- methods for leadership looking forward
- measurement of individual productivity
- When to not measure?
- Enabling a diverse workforce with appropriate tools
- building a workforce which supports disruptive innovation
- alternate data streams in Windows
- principles of enterprise compute architecture
- quantitative, timely, and transparent decision making
- challenge: design processes which involve humans
- failure log
- technical debt
- Conversational openers with a stranger
- what it feels like to have a great leader
- impact of communication on negotiation
- if you want a boulder to roll, place it at the top of a mountain
- how I can generate "the feels" in other people
- Communication channel inequivalence
- characterizing communication as process between entitites
- Emergence at human scale
- I want to get an education in emergence
- combining predator-prey simulations with genetic evolutionary searches
- how to simulate the world
- user specific Hacker News suggestions
- outline for app which provides creative prompts for texting
- common data types useful for data science
- approach for analysis and query of a corpus of documents
- what to measure when measuring my own computer time?
- monitoring myself in Linux: keystrokes, CPU, active windows via bash
- how to evolve an organization, community, or group
- OODA loops and machine learning
- creating an article submitter for Hacker News
- mapreduce
- install hadoop mapreduce on Ubuntu
- how to realistically schedule your time
- mistakes are due to lack of mindfulness
- focusing on dynamic data impacts both data structure and algorithms
- virtualenv for python
- breakdown of submission for the Herox Cognitive Computing Challenge
- brainstorming how to process PDF documents
- how to improve better
- preview of undergraduate degrees
- When to walk away from your customer
- install go on mac os x
- Size of memory limits the complexity of what is teachable
- The value of small talk
- How to solve (and not solve) problems: an example of getting water in a village in Africa
- installing graphviz on Mac OS 10.11
- methodology for people acting as cultural change catalysts
2015 55
- An opportunity for the Cognitive Computing Challenge to succeed - by using objective judgement
- Setting up Oracle Database in Oracle Linux
- set up MySQL on Ubuntu
- human interaction tropes
- provably optimal programs, limits of computation
- failing fast in context of your customer
- reasons the Herox Cognitive Computing Challenge was unsatisfied
- mpi bandwidth as a bottleneck
- perf rapl bottlenecks power performance
- brainstorming transformation tools for text
- my experience with the Herox Cognitive Computing Challenge
- risk associated with errors
- run a python script from a web browser
- challenges associated with "reading" a document
- aspects of tasks
- can the difference between dirty and clean data be quantized?
- top languages I would learn if I had time
- mistake log
- notes for day 4 of the big data workshop
- difference between "big data" and "big compute"
- forecasting as a skill
- notes for day 3 of the big data workshop
- notes for day 2 of big data workshop
- notes from day 1 of big data workshop
- characterizing necessary steps prior to analysis of big data
- big data reading list, coming from MPI/HPC background
- function call graphs across a multitude of software
- Phone interview with Uber for engineer position
- optimal programming team size
- choices on improving a system
- management is not necessarily leadership
- how to minimize the number of transactions among multiple parties paying multiple bills
- configuring a CMS 50D+ pulse oximeter
- PhD in information theory
- notes on bureaucracy and the social network
- notes on leaders and leadership
- meme notes
- planning my next job
- crackpot examples
- A method for herding cats
- how to make a difference
- the gospel of measurement
- altered decision making
- What I learned from Tara Brach
- Practical Joyfulness
- notes on learning C
- open science milestone! NSF
- I don't want to live forever
- Practical Happiness (next is Joyfulness)
- Now what? Transition to emotional development
- problem solving notes from course
- Getting started with Chapel
- Getting started with Android app development on Mac OS X
- two choices are too few as a scientist
- Useful skills not taught in school
2014 13
- how to understand meetings at work
- What I learned from my sister about problem solving
- What are the foundations of how I think?
- Information density of communication
- quantification of memes; the role of emotion
- fomenting revolutions in knowledge
- difference between scientists in academia and government
- I am a scientist
- trouble shooting and problem solving approaches
- Science is not boring
- words and phrases found in the education process
- happiness as a goal
- Why documentation normally sucks, and how to solve the problem
2013 6
2012 48
- how to not be average
- The Actual Scientific Process
- Bashrc
- writing a scientific paper is just telling a story
- how to write a scientific article
- Sabbatical Fellowships for faculty
- social skills are necessary for science researchers
- why blog anonymously in science?
- common problems in presenting scientific data
- computational physics graduate students who change input parameters
- nothing new under the sun?
- importance of workflow in physics
- books on how to be successful scientist
- best practices for computational physics
- how does science research actually work?
- fellowships for graduate school
- teaching approach: graduated introduction
- notes on consulting experience
- best method for data management in research environment
- how to help other people
- how to learn your research topic
- how to actually think different
- what is often not known by new graduate students in computational physics
- how to find a good mentor
- why are physicists unique?
- model for humans
- the role of social engineering in science
- advice from senior scientists
- what career paths are available in physics?
- the importance of failure in making progress
- the art of citations in a scientific paper
- How to write a scientific paper for a peer-reviewed journal
- Math proofs
- how to have a successful interview
- how to be a successful research assistant
- moving to a new job: what questions to think about
- How to program well
- how to search, or, improving your google-fu
- how to fund graduate school
- postdoctoral research positions in science, computational physics
- Licensing your dissertation under Creative Commons
- General observations on scientists
- do computer scientists do science?
- Why should I change schools for college and graduate school?
- Questions to ask during a visit to a prospective graduate school
- Obvious questions in research
- How to succeed as an attendee at conferences
- learning how to learn
2011 35
- my dictionary
- skills necessary for a scientist
- double blind studies of teaching methodology
- Steps in computational research
- Emotion in research
- Prevalence of High-Functioning Autism and Asperger's Syndrome and consequences in science
- How does research lead to new knowledge?
- How to organize a successful conference
- view of PhDs by people without PhDs
- what do smart people do when stuck on a problem?
- Mythical Human month, parallel scaling, and the scientific process
- What makes a good science presentation?
- High Performance Computing as a tool in Science
- Research at a National Lab versus in Academia
- Presentation notes: mistakes to avoid
- why differentiate between information and knowledge?
- thinking scientifically outside the lab
- how to get smarter: mistakes
- how to be a good post doc [links]
- An Academic Career: teaching, research
- Which skills are useful after graduating from graduate school?
- how to get smarter: multi-time scale planning and back-timing
- How is research done?
- financing your science education
- school optimizes you for what?
- science eduction audio/video online
- sharing data in the sciences
- books about physics in general
- scientific approach to teaching
- what there is to know about college and graduate school (after high school)
- compiling external source code from within latex
- latex boolean switch
- latex header
- Why does the scientific method result in progress?
- standing in line