5886
Asking the right question is more important than getting the right answer
The great questions are tractable and fruitful. They lead you on a path of discovery. It is easy to ask how to cure cancer, but that’s not a good question because it does not help anyone do medical research.

Secret questions are the best: if you are the only one with this question in mind, then you may be holding a gold mine. Questions that everyone is having are proportionally worthless.


How might we ask better questions?

1. Pay attention to what is around you and violates your worldview. How did Fleming discover penicillin? He noticed that some mold that had invaded his dirty lab appeared to kill bacteria. He asked the right question at that time.
2. Be patient. Reportedly, Einstein once stated, “It’s Not That I’m so Smart, It’s Just That I Stay with Problems Longer.” The longer you work on a problem, the more likely you are to find interesting questions. (See Forthmann et al. 2018) The easiest way to miss the great questions is to dismiss the problems as uninteresting and move on too quickly.
3. Be physically active, go for a walk. Chaining yourself to a desk is likely counterproductive. I used to think that being an all-out intellectual was the best route, but I now believe that I was grossly mistaken. I personally take a walk outside almost every morning on weekdays. (See Oppezzo and Schwartz, 2014).
4. Don’t be too social. Social pressure toward conformity trigger intense instinctive reactions. It is simply hard to go against the herd. Thus you are better off not know too much about where the herd is. In concrete terms, spend entirely days by yourself.
5. Ask a lot of questions. If you want to become good at providing the right answers, train yourself to answer lots of questions. If you want to become good at asking questions, ask a lot of them.
6. Always question your own thoughts and work.
questions  answers  science  scientificmethod 
45 minutes ago
Outgrowing Advertising: Multimodal Business Models as a Product Strategy
Multi-modal business models - especially for mobile - exemplified by Chinese models. At a minimum, can be used as a basis for inspiration. Creating an ongoing experience.
businessmodels  china  mobile  advertising  subscriptions  gamification  experiences 
6 days ago
When It Comes to Feedback, Start with Yourself
1. Start with Yourself
1a. It’s far easier to change yourself than it is someone else.
1b. Changing yourself leads to change in others.
1c. Your expectations drive other people’s behavior.
2. Giving Critical Feedback.
2a. Don’t be nice, be kind .
2b. Don’t do “The Sandwich”.
2c. Don’t jump into problem solving .
3. Create Space for Giving Feedback
4. Make Feedback Concrete
4a. Don’t personalize
4b. Do not “label” someone 
4c. Do not use “blur” words
5. Connect the Behavior to the Bigger Picture
6. Don’t Make it About Being Right or Wrong
7. Continuously Deliver Feedback
7a. Repeat yourself 
7b. Give feedback early and often
8. Deeper Feedback
9. Limitations of Feedback
9a. Going Beyond Feedback. Provide process to change habits.
10. Giving Praise
feedback  management  leadership  radicalcandor  crucialconversations 
7 days ago
14 Expert Ways To Tell If Clothes Are Well-Made Or Super Cheap
1. To quickly assess an item's quality, hold the fabric up to a bright light.
2. Do the "scrunch test" to see if clothes stay wrinkly.
3. And for a quick quality check, do the "pull test."
4. When in doubt, shop in the men's department.
5. Avoid exposed zippers since they can be a sign of low quality.
6. Check to see if skirts and pants have decent hem allowances — especially if you're tall.
7. And keep in mind that well-made pants have ~French~ seams.
8. Always check the care instructions label for natural fibers.
9. Make sure your pattern matches at the seams.
10. You should also check the seams to make sure all of the thread matches.
11. You can almost always tell the quality of an item by the BUTTONS and the BUTTONHOLES.
12. Gently pull the seams taut to see if there are gaps between the stitches.
13. Linings are always a *good* sign.
14. If you're trying to make sure an item is actually hand-stitched, make sure the seam stitching isn't straight.
quality  clothes  clothing  indications 
7 days ago
Random.org
A random number generator service based on atmospheric noise.
random  generator  numbers  service  free  number 
13 days ago
David Moews' library of public-domain books and papers
Useful collection of PDFs of old public-domain books and papers.
See also, the home page at
http://djm.cc/dmoews.html
pdf  books  articles  publicdomain 
13 days ago
Time for a Change: Introducing irreversible time in economics - Dr Ole Peters
Youtube video of a talk by Ole Peters at Gresham College that talks about the difference between ensemble averages and time averages. Also, history and good examples.

An exploration of the remarkable consequences of using Boltzmann's 1870s probability theory and cutting-edge 20th Century mathematics in economic settings. An understanding of risk, market stability and economic inequality emerges.

The lecture presents two problems from economics: the leverage problem "by how much should an investment be leveraged", and the St Petersburg paradox. Neither can be solved with the concepts of randomness prevalent in economics today. However, owing to 20th-century developments in mathematics these problems have complete formal solutions that agree with our intuition. The theme of risk will feature prominently, presented as a consequence of irreversible time.The solution of the leverage problem is well known to professional gamblers, under the name of the Kelly criterion, famously used by Ed Thorp to solve blackjack. The solution can be phrased in many different ways, in gambling typically in the language of information theory. Peters pointed out that this is an application of the ergodicity problem and has to do with our notion of time.

Transcripts and other information at http://www.gresham.ac.uk/lectures-and-events/time-for-a-change-introducing-irreversible-time-in-economics
olepeters  ergodicity  ensembleprobability  timeprobability  kellycriterion  statistics  time 
14 days ago
5 Lessons Learned From Writing Over 300,000 Lines of Infrastructure Code
Useful wisdom gained from real-world devops.

Lessons:

1. The Production-Grade Infrastructure Checklist.
2. The toolset. Docker, Packer, Terraform.
3. Large modules considered harmful.
4. Infrastructure code without automated tests is broken. Terratest.
5. The release process. Testing, QA, production environments.
devops  checklist  testing  depolyment 
17 days ago
Recent Advances for a Better Understanding of Deep Learning − Part I
Some issues in trying to understand the theory of deep learning.
This article deals with non-convex optimizations (as involved in the loss function), why stochastic gradient descent even works at all (and under what conditions), and the curse of dimensionality.
deeplearning  theory  sgd  stochasticgradientdescent  dimensionality  nonconvex  lossfunctions 
17 days ago
Logic, Explainability and the Future of Understanding
A long, wide-ranging article by Stephen Wolfram. Ostensibly about developing computer-assisted proofs by enumeration. But riffs into philosophical discussions around the nature of doing science, mathematics, and proofs. In particular, the explainability and understandability of proofs by humans. A long deep read.
wolfram  mathematica  logic  theorems  proofs  axioms  science  philosophy  mathematics  understandability  explainability  abstraction  knowledge 
24 days ago
Leak Mitigation Checklist
Procedure to deal with the beginning of your response when you find that information has been leaked.
mitigation  security  leaks  credentials 
26 days ago
Why do you prefer Clojure over Haskell?
Useful discussion over at Clojureverse about Clojure versus Haskell.
Good read.
clojureverse  haskell  clojure 
27 days ago
CIDER's Orchard: The Heart
A three-part series on Cider's Orchard.
cider  orchard  clj  clojure  emacs 
4 weeks ago
Peter Gasston | People don’t change
Youtube video of a presentation at Front-end London, eloquently demonstrating that people have behaved pretty much the same across the ages.
history  future  people  behavior  video  youtube 
5 weeks ago
Convolution arithmetic
A visualization of convolutional arithmetic. Animation of the impact of changing strides and padding.
convolution  convolutionarithmetic  visualization  stride  padding 
5 weeks ago
APPLE’S NEW MAP
Great, long, article comparing the new Apple Maps with the old (and with Google).
maps  directions  apple  google  applemaps  googlemaps 
5 weeks ago
Designing the Ideal Board Meeting
A series of blog posts by Seth Levine on what makes for a good board meeting. Before, during, after. Read them all.
sethlevine  boards  boardmeetings  entrepreneurship 
5 weeks ago
Java Nested Classes: Behind the Scenes
What goes on behind the scene when Java generates nested classes for the JVM.
java  jvm  nestedclass 
6 weeks ago
Execution is Everything
Review of John Doerr's book Measuring What Matters by Tren Griffin.
johndoerr  trengriffin  okr 
6 weeks ago
Lessons learned from creating a rich-text editor with real-time collaboration
Issues involved in building a collaborative editor.
Good read on the issues.
collaboration  texteditor  cscw 
6 weeks ago
Intel Virtualisation: How VT-x, KVM and QEMU Work Together
The several layers at which virtualization works.
Good explanations.
virtualization 
6 weeks ago
Ask the Question, Visualize the Answer
A practical example for how asking and answering questions helps guide you towards more focused data graphics.

Uses the dataset of the projected number of women and men in a population as an example. Then asks several questions (e.g., whether we are interested in absolute values or percentages) and then shows how different visualizations can be sued to answer the different questions on the same dataset.
example  visualization  questions  data  dataset 
6 weeks ago
Stumbling Towards Retirement
1. Construct an internal resume that is about what you value. Only you read it.
2. Only make commitments that are consistent with and further your internal resume. When you do say yes, it should be with full commitment. Be prepared for adverse reactions to you saying no.
3. Do activities that provide the most learning potential, avoid toxic people,,and focus on individuals - not institutions.
4. Re-balance priorities with your spouse.
5. Seek out people you might not otherwise have interacted with. And spend time with them.
6. Every retirement is different.
retirement  priorities  connections  commitments 
6 weeks ago
Why crypto is like the internet (Twitter thread)
Interesting thread on the parallels between the Internet and Blockchain. In particular, the ability for anyone to create a new market in anything. A bit hyped, but worth reading.
internet  cryptography  markets  distribution  economics 
9 weeks ago
Vengeance As Justice: Passages I Highlighted in My Copy of "Eye for an Eye"
Review of "Eye for an Eye" (authored by William Ian Miller).

This is the logic of lex talionis. This is why "an eye for an eye" did not in fact make the whole world go blind. The principle of an eye for an eye, as Miller sees it, is "the more ancient and deeper notion that justice is a matter of restoring balance, achieving equity, determining equivalence, making reparations... getting back to zero, to even." [3] Trading eyes for eyes is not so much about indiscriminate, unthinking violence as it is carefully calculated attempts to match punishment to crime. Talionic justice is a system built on deterrence--not only deterring criminals from committing crimes, but deterring vengeance seekers from exacting too heavy a price in retaliation for crimes committed against them. This is empathy enforced by blood. You think carefully about the pain you inflict on others knowing, that measure for measure, the pain you give others will be given back to you.
justice  vengeance  retaliation 
9 weeks ago
Assessing software engineering candidates
Good read on hiring and evaluating candidates.
Mostly, the idea is to use written material.
Evaluate for the following traits:

- Aptitude
- Education
- Motivation
- Values
- Integrity
interviews  hiring  evaluation 
9 weeks ago
How to Design Programs, Second Edition
Online textbook for teaching programming. A first course on programming. An alternative to SICP. Uses Racket (more specifically DrRacket).
sicp  htdp  teaching  programming  firstcourse  introduction  felleisen  racket 
10 weeks ago
Sequences
Alex Miller's take on Clojure sequences.
Always worth a read.
alexmiller  clojure  sequnces 
10 weeks ago
Inside Clojure's Collection Model
Alex Miller's take on Clojure collections.
Always worth a read.
alexmiller  clojure  collections 
10 weeks ago
Flowchart for choosing the right Clojure type definition form
Clojure offers a number of different forms that define types (and generate Java classes). Choosing between deftype, defrecord, reify, proxy, and gen-class can be a tripping point for those new to Clojure.
clojure  deftype  defrecord  reify  proxy  gen-class  interop  type 
10 weeks ago
The Beauty of Clojure
Clojure component versus mount.
Claims that mount leads to more idiomatic clojure code.
clojure  component  mount 
10 weeks ago
The False Choice Between VC and Bootstrapping
The benefits and challenges of going the VC route. And the same for the bootstrapping route. And the false choice between them. You can do both. Bootstrap until you need financing. And then, choose wisely and on good terms.
Good short article on the tradeoffs.
vcs  bootstrapping  entrepreneurship  startups  financing 
10 weeks ago
Killing the Coding Interview
Instead of asking candidates to code, try:

1. Dig Into Their Experience
2. Have Them Review Your Code
3. Design a System Together

Also, the cases where you should use a coding interview:
a. Determining if entry-level candidates can program at all.
b. If your organization mandates lots of pair programming, pairing in the interview is a good preview of their future work.
c. If a role might require a very specific programming skill and you won’t have time to teach it. This is especially true if you’re on the fence between two seniority levels or titles.
d. When you’re transitioning from a coding interview to a code-free interview.
interview  coding  pairprogramming  questions 
11 weeks ago
Hardening macOS
Some steps to take to harden MacOS.
Worth considering.
macos  hardening  security 
11 weeks ago
Leaked Andreessen Horowitz data reveals how much Silicon Valley startup execs really get paid, from CEOs to Sales VPs
CTO (Sep 2018) median (excluding cash bonuses)
Series A: 240lk
Series B: 250k
Series C: 263k
Series D: 450k

CTO Equity median (consumer)
Series A: 1%
Series B: 1.6%
Sereis C; 1.5%
Series D: 1%
compensation  startups  salary  bonuses  equity 
11 weeks ago
Hardware for Deep Learning. Part 1: Introduction
Semi-regularly updated review of the issues related to deep learning and hardware.

In particular, see Part 3 about GPUs.
hardware  servers  deeplearning  dlcpus  gpus 
11 weeks ago
Why building your own Deep Learning Computer is 10x cheaper than AWS
First of a good series on building your own deep learning machine.
aws  deeplearning  dl  server  host  rig  diy  hardware 
11 weeks ago
How Are Structured Logs Different From Events?
The case for logging structured data. And tons of it (never know what you'll need).

See also the twitter thread
https://mobile.twitter.com/mipsytipsy/status/1042817542648082432
logs  logging  events  structuredata  structure  structuredlogs 
11 weeks ago
Machine Learning Confronts the Elephant in the Room
A visual prank (adding an elephant profile into a living room scene) exposes an Achilles’ heel of computer vision systems: Unlike humans, they can’t do a double take.

When human beings see something unexpected, we do a double take. It’s a common phrase with real cognitive implications — and it explains why neural networks fail when scenes get weird.

The human visual system says, ‘I don’t have right answer yet, so I have to go backwards to see where I might have made an error. Most neural networks lack this ability to go backward.
ml  machinelearning  imagerecognition  failure  adversarialattacks  feedforward  selectivetuning 
11 weeks ago
3 facts about time series forecasting that surprise experienced machine learning practitioners.
1. You need to retrain your model every time you want to generate a new prediction.
2. Sometimes, you have to do away with train/test splits.
3. The uncertainty of the forecast is just as important as, or even more so, than the forecast itself.
timeseries  forecasting  forecasts  ml  machinelearning  uncertainty  retraining 
11 weeks ago
Atlassian Team Health Monitors
Forms and processes to monitor the health of teams. Project teams, leadership teams and service teams. They also provide playbooks to address deficiencies. Essential reading. From Atlassian.
See also
https://www.atlassian.com/team-playbook
atlassian  teams  culture  evaluation  monitoring  monitors  teamhealth  deficiencies 
11 weeks ago
Glitch Employee Handbook
Radically transparent handbook for Glitch employees.
Good reading. Good ideas.
glitch  handbook  employeehandbook  culture  hr 
11 weeks ago
Notes to Myself on Software Engineering
A list of reminders when developing software.
On Development, API design and software careers.
Worth a read.
software  careers  design  api  philosophy 
11 weeks ago
Your Calendrical Fallacy Is...
A list of all the various false beliefs we have about calendars.
calendars  fallacy  fallacies  beliefs  false 
11 weeks ago
Interviewing is a noisy prediction problem
Good, in-depth article on the technical interviewing process. All the things that can go wrong. What some good/bad signals are. Recommends combining various types of signals. And warns against biases. Good read.
interviews  design  process  questions 
11 weeks ago
Measuring the Effectiveness of Error Messages Designed for Novice Programmers
Good error messages are critical for novice programmers. Recognizing this, the DrRacket programming environment provides a series of pedagogically-inspired language subsets with error messages customized to each subset. We apply human-factors research methods to explore the effectiveness of these messages. Unlike existing work in this area, we study messages at a finegrained level by analyzing the edits students make in response to various classes of errors. We present a rubric (which is not language specific) to evaluate student responses, apply it to a course-worth of student lab work, and describe what we have learned about using the rubric effectively. We also discuss some concrete observations on the effectiveness of these messages.

See also https://cs.brown.edu/~sk/Publications/Papers/Published/mfk-measur-effect-error-msg-novice/
errormesages  programminglanguages  novices 
12 weeks ago
From Private To Public: How To Read An S-1
Good article on how to read and understand S-1 filings.

Meet The S-1 Filing
Where To Find S-1 Filings
Financial Nuts and Bolts
Sizing The Offering
Company Details
Risks
Every S-1 Is Special
s1  goingpublic  ipo  sec 
12 weeks ago
The Error Model
A reasonably good exploration of the error models in a programming language and system. Common LISP's condition system is a conspicuous exception. Long.
errormodel  errors  programmingerrors  errorcodes  checkedexceptions  exceptions  preconditions 
12 weeks ago
The seven tools of causal inference with reflections on machine learning
The usual great synopsis by Adrian Colyer at A Morning Paper, of Judea Pearl's paper, on the differences between machine learning models and structural causal models.

See the original paper at
http://ftp.cs.ucla.edu/pub/stat_ser/r481.pdf
causality  inference  interventions  counterfactual  structuralcausalmodels 
12 weeks ago
What-If Tool
Building effective machine learning systems means asking a lot of questions. It's not enough to train a model and walk away. Instead, good practitioners act as detectives, probing to understand their model better.

But answering these kinds of questions isn't easy. Probing "what if" scenarios often means writing custom, one-off code to analyze a specific model. Not only is this process inefficient, it makes it hard for non-programmers to participate in the process of shaping and improving machine learning models. For us, making it easier for a broad set of people to examine, evaluate, and debug machine learning systems is a key concern.

That's why we built the What-If Tool. Built into the open-source TensorBoard web application - a standard part of the TensorFlow platform - the tool allows users to analyze an machine learning model without the need for writing any further code. Given pointers to a TensorFlow model and a dataset, the What-If Tool offers an interactive visual interface for exploring model results.
tensorflow  tf  tersorboard  whatif  ml  machinelearning  tools 
september 2018
How To Be Wrong
Fantastic essay on how to modestly be wrong and why.
If you have a disagreement, run experiments based on the disagreement to see where the truth might be.
By admitting error graciously, you antagonize fewer people and so they are less likely to summarily reject your opinions in a subsequent disagreement.
Make more (preferably small) mistakes. One of the ways of being wrong is "early and often". So, you get used to mistakes, when/where you make them and the areas most likely to lead to mistakes.
If folks have changed opinion from A to B, but none from B to A, then B is more likely correct - even if more people believe in A over B.
Making mistakes often leads to surprising discoveries.
mathematics  mistakes  errors  wrong  heuristics 
september 2018
Learning to Learn
The Art of Doing Science and Engineering: Learning to Learn" was the capstone course by Dr. Richard W. Hamming (1915-1998) for graduate students at the Naval Postgraduate School (NPS) in Monterey, California.
hamming  youtube  learning  science  engineering 
september 2018
The Strange Numbers That Birthed Modern Algebra
The 19th-century discovery of numbers called “quaternions” gave mathematicians a way to describe rotations in space, forever changing physics and math.
quaternions  algebra  3drotations  rotations  numbersystems  methematics  physics  quantummechanics 
september 2018
Forecasting: Principles and Practice
This textbook is intended to provide a comprehensive introduction to forecasting methods and to present enough information about each method for readers to be able to use them sensibly. We don’t attempt to give a thorough discussion of the theoretical details behind each method, although the references at the end of each chapter will fill in many of those details.
ebook  forecasting  statistics 
september 2018
Product Management Mental Models for Everyone
1. Return on Investment.
2. Time value of shipping.
3. Time Horizon.
4. Expected Value.
5. Working Backwards (Inversion).
6. Confidence determines Speed vs. Quality.
7. Solve the Whole Customer Experience.
8. Experiment, Feature, Platform.
9. Feedback Loops.
10. Flywheel (recursive feedback loop).
11. Diminishing Returns.
12. Local Maxima.
13. Version two is a lie.
14. Freeroll.
15. Most value is created after version one.
16. Key Failure Indicator (KFI).
productmanagement  mentalmodels 
august 2018
Financial Modeling for Startups: An Introduction
A somewhat basic financial model for startups.
A good base to build on.
entrepreneurship  startups  finance  financialmodel  excel 
august 2018
The Elephant in the Room
How to fool deep learning.

We showcase a family of common failures of state-of-the art object detectors. These are obtained by replacing image sub-regions by another sub-image that contains a trained object. We call this "object transplanting". Modifying an image in this manner is shown to have a non-local impact on object detection. Slight changes in object position can affect its identity according to an object detector as well as that of other objects in the image. We provide some analysis and suggest possible reasons for the reported phenomena.
deeplearning  computervision  objectrecognition  failures  recognitionfailures 
august 2018
How Technology Grows (a restatement of definite optimism)
Technology should be understood in three distinct forms: as processes embedded into tools (like pots, pans, and stoves); explicit instructions (like recipes); and as process knowledge, or what we can also refer to as tacit knowledge, know-how, and technical experience. Process knowledge is the kind of knowledge that’s hard to write down as an instruction. You can give someone a well-equipped kitchen and an extraordinarily detailed recipe, but unless he already has some cooking experience, we shouldn’t expect him to prepare a great dish.

- Process knowledge is represented by an experienced workforce.
- The US industrial base has been in decline.
- Knowledge should circulate throughout the supply chain, flowing both up and down the stack.
- Let’s try to preserve process knowledge.
- The future should be more than services.
- How are we going to do science fiction without an industrial base?
- The US should emulate a different country. (Germany).
- Development is the only hard truth (or, the social consequences of economic growth).
- Optimism as a propellant of growth (or, more industry and less Twitter).
- People with the right priorities. Bill Gates, Freeman Dyson, Andy Grove.
- Does better capital allocation always lead to technological acceleration?
- What are some challenges to these views?
- Definite optimism as human capital.
technology  tools  skills  processes  tacitknowledge  optimism 
august 2018
The North Sense
The North Sense is an exo-sense intelligently designed for evolution, which means it sits outside the body but is permanently attached. It allows a person to sense the electromagnetic field of the planet. The North Sense is typically attached to the upper chest and gently vibrates when facing magnetic north.
senses  exosense  cyborg 
august 2018
How to find stuff in Git
Good examples of commands in git to locate different files, lines of code, changes, authors, etc.
Useful.

Also at https://www.tygertec.com/find-stuff-git/
git  examples  find  locate 
august 2018
Utilizing a Rubric to share expectations of the QA Engineer Role
Good example of using rubrics (method to evaluate performance) for different roles. In this case, for QA. Mostly good as an exemplar.
example  rubrics  roles 
august 2018
Parameterize, execute, and analyze notebooks
What the title says.
Treats notebooks as immutable functions and schedules applying specified inputs to the notebooks to generate outputs.
notebooks  scheduling  paramterization 
august 2018
Bumper-sticker Computer Science
Pithy CS phrases. From Jon Bentley of Programming Pearls fame.
Wisdom.
wisdom  computerscience  sayings  phrases 
august 2018
Stayzilla will reboot its operations
Why Stayzilla failed. And lessons learned. What metrics might not be as useful.
marketplaces  stayzilla  metrics 
august 2018
« earlier      
advertising advice adwords agile ai algorithm algorithms amazon analysis analytics api architecture awesome aws bigdata bitcoin book books brain browser business cache career charts cheatsheet clojure cloud code communication comparison complexity concurrency course creativity cryptography css culture data database datascience deeplearning design development devops distributed diy ec2 economics education encryption engineering entrepreneurship errors evolution exchange explanation facebook failure filetype:pdf finance fonts food framework free functional future git github google graphics guide hardware health hiring history howto html humor information innovation inspiration internet interview interviews introduction java javascript js jvm keyword keywords language languages law leadership learning legal lessons library life lifehacker links linux lisp list lists local machinelearning management maps marketing marketplaces math mathematics media:document memory metrics microsoft mistakes ml mobile money monitor monitoring music mysql negotiation network networking neuroscience online opensource optimization papers patterns pdf performance philosophy physics politics ppc presentation presentations pricing privacy probability problems process productivity programming psychology python quality questions reading reference research resource resources review reviews sales scalability scaling science scm search security sem seo server shopping sleep social software speed startup startups statistics storage strategy sysadmin teaching teams technology ted test testing theory thinking tips tool tools travel tutorial tutorials typography ui usability ux value vc video visualization web webdesign webdev windows work writing youtube

Copy this bookmark:



description:


tags: