Extra time

Seth Lloyd

I watched some videos about Seth Lloyd about quantum computers and quantum machine learning before. It is always good to know that there is something beyond what we think we already know.

I had never thought about how the evolution process took place in our communication media. It is something that we use every day and that we normalize like text messages, social networks and the now new information transfer networks, but what is there beyond? Where does it come from?

It is interesting how millions of atoms and particles of elements that surround us not only exist because of a “standard” configuration. But even these as it comes to raise can contain information stored as a result of the connection established with its neighbors.

In how they interact, this is exemplified with a machine that shoots particles and that form a pattern determined by a wave, like what we learned last week. The information contained within the particles influences the behavior of others, that’s why that phenomenon is called interference pattern. And what is specific about how our genes interact and, despite coming from the same father and mother, mutations and changes are generated in our genes as a result of many values ​​that until now can’t be calculated with precision,, but it is good to know that they are there, because at some point we’ll find a way to access that information.

Why should we have our own black box?

This is something I have been listening to a lot recently, “Fail fast to learn fast” I can’t be more agreed, why black boxes? Most people probably know about Airplane’s black boxes, these boxes have records for both pilots and systems so in case of an accident there can be clues about what went wrong and how to prevent it from happening again, thanks to that most of the time the same accident won’t happen again. Another more modern and famous proof of this are the spaceX spacecraft, I remembered seeing the grasshopper’s test, I loved seeing it going further and further into space, then all the tests of the falcon heavy landing which most of the time ended in explosion until they finally succeed is very inspirational. If there is an example of black boxes and learning from failure, for me is this.

Richard Feynman the great explainer great minds

“Not all great minds are great explainers” First things first, I was curious about the bongos so I looked about it and found a video which was very funny. I usually don’t look about this kind of thing but because it was a regular gag I have to. Now to the point, Richard was curious about everything, so even when he didn’t have access to an education which could fill his curiosity he took all the books he could and started learning on his own.

Stephen wolfram computing theory of everything

the idea of computation, we must know how to use all the computational resources out there for human purposes. One of the things I like about programming is putting ideas into code and making it work, but sometimes we spend a lot of time trying to solve something ourselves that we don’t look at all the computational resources out there that can help to build something better.

Something interesting about these videos aside from the idea of having a new vision of problems was the fact that Stephen wanted to have an interpretation about the universe. Looking at some of the models he showed I noticed one that was big news last year, the most modern interpretation of what the universe is made of. The video is from 2010 so seeing one of his dreams getting closed is amazing.

There is a relationship between Seth Lloyd and Stephen Wolfram. They think of computing as a study tool. Despite being a species with “superior” cognitive abilities in aspects such as logic, mathematics, etc. We have certain limitations since we have reached a point of study in which the amount of data to process is too big. This is why quantum computers and machine learning are becoming more necessary.

Feynman on scientific method

No matter how smart you are when you are looking for a law, if the experiment disagrees with your guess is wrong. We can only be sure we’re wrong, the experiment that proves to be right today can be proved to be wrong tomorrow.

Pretotyping manifesto

The pretotyping manifesto

Innovators beat ideas

Pretotypes bead productypes

Doing beats talking

Simple beats complex

Now beats later

Commitment beats committees

Data beats opinions

Don’t look for ideas but innovation.

Pretotyping. Validating the market appeal and actual usage of a potential new product by simulating its core experience with the smallest possible investment of time and money.

Pretotyping vs Prototyping

Pretotyping is about:

Investment: hours, days

Main question: would we use it?

Deliverable: [working] pretotype

Advantages: You get to fail fast

Prototyping is about:

Investment: days, weeks

Main question: can we build it?

Deliverable: Working prototype

The more time you spend on an idea the harder is to admit failure

Pretotyping example: Mobile apps

Aim for data instead of opinions

I’ll take some space here to expose my team’s idea for pretotyping

1. MOBILE APP FOR ADOPTING PETS

Idea description

This idea consists of a mobile application where users are able to put their pets up for adoption. A user looking for a pet to adopt could browse a list of different photos of pets that their owners want to give up for adoption. A user selects a pet that she likes and the app shows the pet description and a way to contact the owner. The goal is to connect animals looking for a family and people who are looking for companionship. The mobile app is called Pet Me.

Hypothesis

There are a lot of mobile apps where you can buy and sell products, find a significant other, or enroll is an online courser. The basic principle is this: meet offer with demand. In this way, a seller finds a buyer, people who want to date find each other, a student finds a teacher, etc. The pet adoption app is a remix thereof. The functionality would be pretty much the same, just the type of service changes. The hypothesis is that there are a lot of people who want to adopt a pet and there are a lot of people who need to give a pet up for adoption. The goal of the app is to connect the two.

Experiment description

The experiment consists of a survey where people can rate and give their opinion about the pet adoption app. The survey contains the link to the pretotype built with the tools found in https://www.figma.com/. This pretotype is a set of high-fidelity wireframes that simulates a real mobile app the user can interact with. After the user is finished checking the pretotype, she answers the five questions of the survey.

Validation

The pet adoption app will be implemented if the results of the survey show the following:

• The app is rated with a minimum of three stars.
• The NPS in the second or third questions are greater than zero.
• At least half of the people who answered the survey considered the app useful or fun.
• At least half of the comments in the fifth question are positive.

The pet adoption app will be discarded in any other case.

Results

These are some of the comments people wrote. Some comments are negative, but most of them are positive.

2. APP FOR SECURITY SOCIAL NETWORK

Sometimes people publish information that can be used to steal the identity of that person and most of the time the person doesn’t know that the publication has content that can give information about their physical condition or where is his current location. For this is designed an application that is capable of analyzing the text and images that are published and warn that person that it is necessary to edit the publication so this information cannot be used in unwanted activities.

Hypothesis

People increasingly like to know that the internet has less information about oneself so the application will receive good acceptance from users that warns them that the publication contains sensitive information such as the user’s location or where they live. The application will be very useful for users who use more than one social network to which they make daily publications.

Experimentation Description

Users will see part of how the application would be like and will be asked questions about their opinion of the application and if they like it. Also, we ask if they have had problems with any publication they have made in a social network and if they are interest in an application that would prevent publish that information. In the survey, you can see the design of the application, only you need to open a website with the link in the survey and we publish a link to the survey of microsoft forms in different social networks and obtain the opinion of different users.

Each question will help us to make changes to the application and see if the user is satisfied with the usefulness of the application and we can observe the acceptance of the idea by users with this information.

Validation

For the validation of the application, the results of the survey will be taken to decide how design the application and the app will be implemented according to the following results:

• Most of the users had or maybe had a related problem in a social network.
• More than half of the users would like an application that warns them of possible security problems.
• If users would recommend the application to other people with a rating greater than 7.

If these parameters are not passed, the application will not be developed.

Results

3. Web Preview

Web Preview is an application that intents to get a fast screenshot/preview of the website before opening it. This is focuses mainly for shortened links or untrusted websites.

Hypothesis

When searching for something there will be times where the user will end up finding a website that seems to have the information, they are looking for but in the end is just misinformation or not what the user expected. Other times they will be asked to be redirected to an unknown site. Another common scenario is when they just don’t trust a shortened link or website.

Experiment description

The application works like YouTube video’s preview, the user can’t interact with the result, is just a static image of a website so the user can be sure the website has what they are looking for. Simple version

Validation

We are trying to discover if people would actually use this application based on the following statements.

• How often they get the wrong information
• How they feel about spending time looking at a screenshot
• If they would actually like it to be and application

Results

We got positive results about the implementation of a Web Preview.

We got mixed results about implementing it for shortened links and also for the fact that the application could make the user waste more time when looking for information.

The most interesting result is the implementation of the idea for Browsers instead of an application, which was in the original scope of the idea.

Testing and automation

Nowadays almost anyone has access to devices such as smartphones, computers, smartwatch. With access to various work platforms such as Google, Netflix or Amazon. We are aware of the number of people who consume their products, but not much about the management of their processes. More specific for this case and as a weekly topic. We do not know how they can carry out the tests to their systems, codes, functions or applications.

As the first point in common between google and Netflix is this culture of testing. Putting everything to the test at different times of production. During the design and with good planning it helps them create a system that could be called robust, not fail-safe and that allows them to optimize quickly.

One of the most interesting topics is parallel testing. While testing a product it seems that they are running tests, if something happens and changes are made, another test runs from that new point with the changes made previously. The first thing that I thought and that was mentioned in the video is that it is a process that is quite expensive speaking of computational resources, but it is still necessary.

Another important fact is that they see the testing culture as everyone’s duty to create a code that is not only functional but also optimized. It makes it easier for the person in charge of that code and who knows what he is doing to dedicate himself to delivering it as well as possible from the beginning, if it is accompanied by some tests, the better. It is a continuous collaboration.

They handle it from the design, taking into account the needs or problems that users already have. The strategy is to know how the user is going to consume the product.

They’ll find restrictions like those that exist in iOS devices, when they wanted to apply tests on users, but that option was not viable for security reasons. For this, other approaches are used such as automated testing tools or later it is commented that a limited number of users are released so it could be added to test the applications.

At this point I imagined that all the tests were better to use them on applications in a safe environment, but what happens when you cannot? Like Netflix with such a huge streaming service. The creation of tools to experiment on a production system to build confidence by subjecting it to future modifications.

As I mentioned at the beginning, purpose and design with the purpose of avoiding failure is not enough. There are so many variables that inevitably something will go wrong along the way, because as they say, better be prepared for that moment.

Something important needed during the testing phase is that we need to develop a better ability to report a bug. We had already seen something similar in which we considered how to interact with a busy boss when we need to ask him for something.

• Report the failure.
• Posibles causas.
• How to fix it.
• Make the request.

Recap

This week assignment was interesting, pretotyping is an interesting technique that allow us to test our ideas in a short range of time, the fastest we fail the fastest we improve. From the Feynman’s scientific method we must always asume we are wrong so we can think of ways to improve.

The testing and automation made me feel a bit uncomfortable, what I mean is that if I haven’t watched those videos the idea of doing test in production couldn’t have even cross my mind. About breaking things at Netflix, it make sense doing it at college with our project but doing it with a services with millions of users sounds a bit scary but the fact that the service is always available means that they know what they are doing and its working.