Notes from 4–10/11

Are You a Moral Grandstander?

Do you strongly agree with the following statements?

• When I share my moral/political beliefs, I do so to show people who disagree with me that I am better than them.
• I share my moral/political beliefs to make people who disagree with me feel bad.
• When I share my moral/political beliefs, I do so in the hopes that people different than me will feel ashamed of their beliefs.

1. Moral grandstanders (those scoring high on the moral grandstanding survey) tend to also score high in narcissistic characteristics and also tend to report status-seeking as their fundamental social motive.
2. There is no relationship between moral grandstanding and political affiliation. However there is a link between moral grandstanding and political polarization: people on the far left and far right are both more likely to score higher in moral grandstanding characteristics than those who are more moderate democrats and republicans.
3. Moral grandstanders are more likely to report greater moral and political conflict in their daily lives (e.g., “I lost friends because of my political/moral beliefs”) and they report getting into more fights with others on social media because of their political or moral beliefs. This correlation was found even after controlling for other personality traits, and continued over the course of a one-month longitudinal study.
4. Grandstanders were more likely to report antagonistic behavior over time, such as attacking others online, or trying to publicly shame someone online because they held a different moral or political belief.

• The need to seek social status by dominating others (“When I share my moral/political beliefs, I do so to show people who disagree with me that I am better than them”)
• The need to seek status through being a knowledgeable and virtuous example (“I want to be on the right side of history about moral/political issues”, “If I don’t share my views, others will be less likely to learn the truth about moral/political matters”, “I often share my moral/political beliefs in the hope of inspiring people to be more passionate about their beliefs.”)

I really hope I don’t come across as this way. Thank goodness I don’t express my morals through words but seek to do them through action.

AI

Neural ODEs: breakdown of another deep learning breakthrough

self.net = nn.Sequential(
            nn.Linear(2, 50),
            nn.Tanh(),
            nn.Linear(50, 2),
        )
true_A = torch.tensor([[-0.1, 2.0], [-2.0, -0.1]])class Lambda(nn.Module):
    def forward(self, t, y):
        return torch.mm(y, true_A)
Spiral function
true_A = torch.tensor([[-0.1, 2.0], [-2.0, -0.1]])class Lambda(nn.Module):
    def forward(self, t, y):
        return torch.mm(y, true_A

Models the dynamics between two consequent dynamic steps, then optimise through the ODE solver that minimises the difference bewteen actual and modeled dynamics.

so this works if what you’re trying to find is some sort of differentiable function.

Which is alot of things

Common Data Mistakes to Avoid

I fell into a trap — this was an advert.

Data

Deep Dive AUC

• gives an idea of how well separated the classes are
• not sensitive to prevalence (ratio between classes)
• shape of curve expresses sensitivity (further to left would be more sensitive)

Confusion Matrix — Deep Dive

Product

OUTCOME BASED ROADMAPS : Unleash the Power of a Shared Vision and Purpose

An Outcome is a measurable change in customer behavior.

Step 1: Start with a clear and compelling Product Vision. Make it clear why your product should exist.

Step 2: Create 2 to 3 Goals for your company for the next 1–2 years that, if achieved, would result in your realizing part of your product vision. Goals should be specific and measurable (ideally SMART goals)

Step 3 is to focus on which customer behaviors you need to change in order to achieve your goal — your Outcomes. These could be things like signup rate, conversion rate, customer satisfaction, items created etc.

Step 4: Identify Opportunities for changing customer behavior. Opportunities are:

• Customer Problems,
• Customer Gains,
• Customer Pains, or;
• Jobs To Be Done

Tools

Learn Python by Building 5 Games

Kinda useful?

Tableau-like Drag and Drop GUI Visualization in R

This is absolutely amazing