The Curious Engineer

All blogs / AI / ML

A Computer Scientist in a Business School

behind-the-enemy-lines.com

Random thoughts of a computer scientist who is working behind the enemy lines; and lately turned into a double agent.

AI / MLScienceEssays / personal

’10’15’20’25
Posts
258 from 8 May 2007 to 22 Jul 2026
Posting
monthly (7 in the last year)
Kind
personal blog
Feed
http://feeds.feedburner.com/AComputerScientistInABusinessSchool/

2026 6

  1. The Lawyer I Never Hired: How ChatGPT Got Me $4,760 Out of Norse Atlantic
  2. Taste Is Not Enough. Reality or Bust
  3. How I Stopped Being a Copy-Paster for My AI Agent: Claude Code, Google Cloud, and the Loop to Close
  4. "Let's Work on the Next Task": Claude Code, GitHub, and the Most Diligent Project Manager I've Ever Had
  5. Listening to My Students at Scale: Exit Tickets, NotebookLM, and the Tightest Feedback Loop I've Ever Built
  6. Everybody Is a CEO Now (And What Exactly Am I Doing Here?)

2025 3

  1. Fighting Fire with Fire: Scalable Personalized Oral Exams with an ElevenLabs Voice AI Agent
  2. Training LLaMA using LibGen: Hack, a Theft, or Just Fair Use?
  3. Copyright, Fair Use, and AI Training

2024 3

  1. Developing Grading Rubrics using Docent
  2. Grading with AI: Introducing Docent
  3. The PiP-AUC score for research productivity: A somewhat new metric for paper citations and number of papers

2022 1

  1. Tell these fucking colonels to get this fucking economist out of jail.

2021 1

  1. "Geographic Footprint of an Agent" or one of my favorite data science interview questions

2019 1

  1. Mechanical Turk, 97 cents per hour, and common reporting biases

2018 2

  1. Distribution of paper citations over time
  2. How many Mechanical Turk workers are there?

2017 1

  1. Why was my Amazon Mechanical Turk registration denied?

2016 3

  1. AlphaGo, Beat the Machine, and the Unknown Unknowns
  2. A Cohort Analysis of Mechanical Turk Requesters
  3. The Decline of Amazon Mechanical Turk

2015 4

  1. An API for MTurk Demographics
  2. Postdoc Position for Quality Control in Crowdsourcing
  3. The World Bank Report on Online Labor
  4. Demographics of Mechanical Turk: Now Live! (April 2015 edition)

2014 4

  1. My Peer Grading Scheme
  2. Online Markets: Selling products vs. selling time
  3. Future of Education: Fighting Obesity or Fighting Hunger?
  4. Crowdsourcing research: What is really new?

2013 10

  1. Badges and the Lake Wobegon effect
  2. CrowdScale workshop at HCOMP 2013
  3. Online labor markets: Why they can't scale and the crowdsourcing solution.
  4. Crowdsourcing and information theory: The disconnect
  5. Facebook implements brand safety, doing it "manually" (crowdsourcing?)
  6. Mechanical Turk account verification: Why Amazon disables so many accounts
  7. Project Troia: Quality Assurance in Crowdsourcing
  8. Intrade Archive: Data for Posterity
  9. WikiSynonyms: Find synonyms using Wikipedia redirects
  10. Towards a Market for Intelligence

2012 30

  1. Mechanical Turk changing the defaults: The game has changed
  2. How big is Mechanical Turk?
  3. Why I love crowdsourcing (the concept) and hate crowdsourcing (the term)
  4. New version of Get-Another-Label available
  5. Why oDesk has no scammers
  6. Using oDesk for microtasks
  7. The disintermediation of the firm: The feature belongs to individuals
  8. On Retention Rates
  9. Why is oDesk called oDesk?
  10. Discussion on Disintermediating a Labor Channel
  11. Disintermediating a Labor Channel: Does it Make Sense?
  12. Visualizations of the oDesk "oConomy"
  13. The oDesk Flower: Playing with Visualizations
  14. The Emergence of Teams in Online Work
  15. TREC 2012 Crowdsourcing Track
  16. ACM EC 2012 schedule
  17. The Google attack: How I attacked myself using Google Spreadsheets and I ramped up a $1000 bandwidth bill
  18. Feedback, Unemployment, and Crowdsourcing: A Modest Proposal
  19. Philippines: The country that never sleeps (or, When is the world working? The oDesk Edition)
  20. Mechanical Turk: More SETI@Home and less Amazon Web Services
  21. ACM EC 2012 Workshops
  22. The (Unofficial) NIST Definition of Crowdsourcing
  23. When do reviewers submit their reviews? (ACM EC 2012 version)
  24. 50% of the online ads are never seen
  25. Crowdsourcing and the end of job interviews
  26. The Need for Standardization in Crowdsourcing
  27. Mechanical Turk vs oDesk: My experiences
  28. ACM EC 2012: Some early statistics
  29. Identify Verification (and how to bypass it)
  30. HCOMP 2012: 4th Human Computation Workshop

2011 29

  1. Does lack of reputation help the crowdsourcing industry?
  2. BusinessWeek...
  3. Collective Intelligence 2012: Deadline November 4, 2011
  4. Probabilities and MTurk Executives: A Troubled Story
  5. The impact of online reviews: An annotated bibliography
  6. Native vs Grapevine Reputation on MTurk
  7. A tale about parking
  8. Why I will never pursue cheating again
  9. Extreme value theory 101, or Newsweek researching minimum wage on Mechanical Turk
  10. Accepted papers for the 3rd Human Computation Workshop (HCOMP 2011)
  11. Crowdsourcing and the discovery of a hidden treasure
  12. Crowdsourcing Education
  13. Pay Enough or Don't Pay at All
  14. Crowdsourcing: Lessons from Henry Ford
  15. Deadline for HCOMP 2011 extended: Submission due on April 29th
  16. Video from NYC Crowdsourcing Meetup
  17. NYC Crowdsourcing Meetup: April 13th, 6.30pm
  18. Tutorial on Crowdsourcing and Human Computation
  19. An ingenious application of crowdsourcing: Fix reviews' grammar, improve sales
  20. Crowdsourcing goes professional: The rise of the verticals
  21. Uncovering an advertising fraud scheme. Or "the Internet is for porn"
  22. Do Mechanical Turk workers lie about their location?
  23. The Road to Serfdom, ACM Edition
  24. The promise and fear of an assembly line for knowledge work
  25. What was the main factor for Watson's success? Hardware, software, or data?
  26. Browsers of Mechanical Turk workers
  27. 3rd Human Computation Workshop (HCOMP 2011), San Francisco, August 7 or 8
  28. Notes from "Crowdsourcing in Search and Data Mining" (CSDM) workshop
  29. The unreasonable effectiveness of simplicity

2010 40

  1. Amazon Reacts: Spammers Kicked Out of MTurk!
  2. Mechanical Turk: Now with 40.92% spam.
  3. Sharing code, API's, and a Readability API
  4. Excerpts from "The Communist Manifesto"
  5. Wisdom of the Crowds: When do we need Independence?
  6. Mechanical Turk, "Interesting Tasks," and Cognitive Dissonance
  7. NYC, I Love You(r Data)
  8. Introductory Research Course: Replicate a Paper
  9. Cease and desist...
  10. Student websites
  11. Can Crowdsourcing Scale? The Role of Active Learning
  12. A Plea to Amazon: Fix Mechanical Turk!
  13. Mechanical Turk and Data Driven Journalism: The Case of ProPublica
  14. Be a Top Mechanical Turk Worker: You Need $5 and 5 Minutes
  15. The Explosion of Micro-Crowdsourcing Services
  16. The Explosion of Crowdsourcing Workshops
  17. Mechanical Turk Requester Activity: The Insignificance of the Long Tail
  18. What Tasks Are Posted on Mechanical Turk?
  19. Worker Evaluation in Crowdsourcing: Gold Data or Multiple Workers?
  20. Analytics for Class Lectures
  21. Reputation Integration and the Future of Reviews
  22. Mechanical Turk, Low Wages, and the Market for Lemons
  23. Peer Reviewing for Oral Presentations?
  24. Liveblogging from HCOMP 2010
  25. Detecting Spammers on Mechanical Turk, Part II
  26. Detecting Spammers on Mechanical Turk, Part I
  27. Accepted Papers for HCOMP 2010
  28. Prediction Optimizers
  29. Google Prediction API: Commoditization of Large-Scale Machine Learning?
  30. Google Scholar now Supports Email Alerts
  31. Crowdsourcing: Not just a cost saver
  32. KDD Accepted Papers, Deadlines for HCOMP 2010 and SNAKDD 2010 workshops
  33. Yahoo!'s Key Scientific Challenges: Your student is a winner!
  34. Stop Publishing!
  35. Getting High Quality Results on MTurk
  36. Citation Tracker: Now with an API
  37. The New Demographics of Mechanical Turk
  38. Why Mechanical Turk Allows Only US-based Requesters?
  39. Universities and Intellectual Property: A Minefield?
  40. Did you find this helpful?

2009 41

  1. Prisoner's Dilemma and Mechanical Turk
  2. Anchoring and Mechanical Turk
  3. Ignore.. (Test)
  4. Using the NYC Data Mine for an Intro Database Assignment
  5. Utility of Money and the St. Petersburg Paradox
  6. What is the (Real) Cost of Open Access?
  7. When Noise is Your Friend: Smoothed Analysis
  8. Rationality, P=NP, Prediction Markets, and a Paradox
  9. Citation Tracker: Monitoring Citations to your Publications
  10. The different attitudes of computer scientists and economists
  11. Workshop on Information in Networks (WIN)
  12. Get a Consent Form (for IRB) on MTurk using Qualification Tests
  13. Top Requesters on Mechanical Turk
  14. When to Post Tasks on Mechanical Turk?
  15. Workshops: Official or Unofficial Proceedings?
  16. Is Amazon Mechanical Turk a black market?
  17. How Prices Evolve in Prediction Markets?
  18. Books, Journals, Conferences, Blogs
  19. LiveBlogging from HCOMP 2009
  20. Google Fusion Tables: Databases on the Cloud
  21. Acceptance Rate: 100%
  22. Evaluation Feedback and Stakhanovist Research Profiles
  23. Public Paper Reviews
  24. Google App Engine and Java: First Impressions
  25. LiveOps and Human Computation
  26. Fitting a Power-Law with Censored Data
  27. Turker Demographics vs Internet Demographics
  28. Mechanical Turk: Profitable or Not?
  29. Human Computation Workshop (HCOMP 2009)
  30. Positive or Negative? The Economist, Prediction Markets, and Mechanical Turk
  31. Why Failing to Fail is a Failure (Prediction Market edition)
  32. Luis von Blog
  33. Your estimated waiting time: Infinite!
  34. Duration of Tasks on Mechanical Turk
  35. Monitoring the Dynamics of Mechanical Turk
  36. Is it unethical to pay for reviews?
  37. Time Out New York and Mechanical Turk
  38. Mechanical Turk, Human Subjects, and IRB's
  39. How do you charge the Mechanical Turk expenses?
  40. Soliciting Reviews on Mechanical Turk
  41. How good are you, Turker?

2008 47

  1. The Faces of Mechanical Turk
  2. Social Annotation of the NYT Corpus?
  3. Use of Excel-generated HTML Considered Harmful
  4. Academic Trees
  5. Computing State Correlations in Elections Using (Only) Prediction Markets
  6. The New York Times Annotated Corpus
  7. Modeling Volatility in Prediction Markets, Part II
  8. Modeling Volatility in Prediction Markets, Part I
  9. Student websites
  10. Reviewing the Reviewers
  11. VP Debate and Prediction Market Volatility
  12. Sarah Palin and Markov Models
  13. How Much Turking Pays?
  14. Why People Participate on Mechanical Turk, Now Tabulated
  15. Links to Mechanical Turk Articles
  16. Mechanical Turk worker quality and HIT difficulty
  17. Mechanical Turk Allows Bulk Submissions Using Templates
  18. JDMR: Officially live
  19. Using The New York Times Reader
  20. Options Markets for Prediction Markets?
  21. Massive Data and the End of the Scientific Method
  22. Statistical Significance of Sequential Comparisons
  23. The (Statistical) Significance of the Impact Factor
  24. Experimental Repeatability or simply Open Source?
  25. SCECR 2008: Symposium on Statistical Challenges in Electronic Commerce Research
  26. How Much a Paper Submission Costs?
  27. Journal of Database Management Research (JDMR)
  28. The Importance of "Everything Else Being Equal"
  29. Mechanical Turk: The Demographics
  30. Why People Participate on Mechanical Turk?
  31. Liveblogging from Dagstuhl: Day 4 (March 13)
  32. The "Good Movie" Talk and the "Anytime" Talk
  33. Liveblogging from Dagstuhl: Day 2 (Mar 11)
  34. Liveblogging from Dagstuhl: Day 1 (Mar 10)
  35. Mechanical Turk: The Foundations
  36. Course Evaluations and Prediction Markets: The Results
  37. Search and the New Economy: Slides and Videos
  38. Course Evaluations and Prediction Markets
  39. Privacy and Intellectual Property: Two Sides of the Same Coin?
  40. Mechanical Turk, Charity, and Incentives
  41. How to Embed a YouTube Video in PowerPoint 2007
  42. Google-ridiculous
  43. Who is Reviewing the Reviewers?
  44. Misunderstandings of Power-Law Distributions
  45. Are you a Bayesian or a Frequentist? (Or Bayesian Statistics 101)
  46. Are you a Bayesian or a Frequentist?
  47. Definining Probability in Prediction Markets

2007 32

  1. Freakonomized!
  2. Love, Facebook, and xkcd
  3. Political Prediction Markets: Some Thoughts
  4. By Popular Demand: Mitt Romney
  5. Prediction Markets are NOT Efficient
  6. My First Mechanical Turk Paper
  7. GMail Supports IMAP (but forgets to notify the abuse team)
  8. New Class: Search and the New Economy
  9. Only for Database Geeks, the SeQueL
  10. What is Wrong with the ACM Typesetting Process?
  11. Visualizing the Dirichlet
  12. Using Facebook ... as a facebook
  13. Only for Database Geeks
  14. Implicit and Explicit Changes in Contracts: Rent-A-Coder
  15. The Price of Privacy (comScore edition)
  16. Ambiguous First Names and Disambiguation
  17. Experiences using Amazon Mechanical Turk
  18. Experimental Repeatability Requirements for SIGMOD 2008
  19. Only in Greece...
  20. Being Accurate when having an Agenda
  21. Evaluating Information Extraction using xROC Curves
  22. Blending Mind Maps with Wikipedia
  23. Data Cleaning Service
  24. Wikipedia-based Web Services
  25. Brain Activity and Economic Judgment?
  26. Are Publishers Making Themselves Useless?
  27. Are Prediction Markets Efficient?
  28. Playing with Wikipedia
  29. Uses and Abuses of Student Ratings
  30. Impact Factors and Other Metrics for Faculty Evaluations
  31. Names (or Parents?) Make a Difference
  32. Replacing Survey Articles with Wikis?