#9 - Social Status In An AI World
On GPT-3, Warren Buffett, Job Interviews, Social Status and Predictive Coding
This is a weekly newsletter on interesting articles covering business, finance, technology and human behavior. Subscribe by clicking the link below and we will send The Curious Cat to you every Saturday morning directly to your inbox
If you continue receiving this email in your Promotions tab, please move this email in your Inbox and mark it with a star. This would ensure that you get the newsletter in your inbox.
Hello,
Greetings from The Curious Cat.
This week, we have a mixed bag of articles from diverse fields for our curious readers.
In today’s issue we look at -
Decoding Social Status
Understand Warren Buffett’s contribution
Tips to brainstorm in a remote working setup
What is GPT-3?
The Flaws of Job Interviews
Decoding Social Status
Read the article here
There are a set of behavior patterns that persist within humans due to collective form of human nature that helps in decoding social status. The preliminary wisdom on social status lies in learning to distinguish between dominance and prestige.
Dominance is a status we derive from intimidation and prestige is a status we derive from doing impressive things or having impressive traits/skills. These different systems can be broken down into a high status-low status complementary pair where dominance & submission is one pair and prestige & admiration is the other pair.
Both systems are complete opposites:
Avoidance vs Approach - Dominance works by inspiring fear and other avoidance instincts. Prestige inspires admiration
Taking vs Giving - As Adam Grant would say, perks of dominance are taken by force and perks of prestige are given freely
Entitlement vs Gratitude - Dominant individuals expect deference and prestigious individuals show humility
Admiration is active deference to the prestigious individual and hence admiration is prestige seeking. The relationship between an admirer and a prestigious individual is similar to the relationship between an apprentice and a master. Prestige seeking and deference are complementary teaming instincts.
Social status helps in understanding various dynamics of group behavior, individual behavior in a group and encourages to build influence. It can also explain the recent social phenomenon on cancel culture and tribal thinking when you look at them as status-seeking activities.
Does understanding social status help in an environment dominated by data and algorithms? Can artificial intelligence get the necessary prestige from humans?
Buffett’s Eight Contributions Beyond Shareholder Returns
Read the article here
There are a few important lessons from Warren Buffett’s long term track record that need to be understood:-
Educating - A consistent drumbeat of buying long term good businesses at reasonable prices by controlling your temperament and believing in the US stock market over time.
Promoting the importance of controlling emotions and combating biases
Being a long term optimist
Buying stocks when they go on sale
Evolving investment process - Buffett had great success as a value investor in his early days but his pivot to high quality investing is impressive for all.
Focus on fees - Buffett has been openly critical of investment firms and products that charge exorbitant fees.
Respecting simplicity - A large purchase, a firm with long term earnings power, good returns on equity, little debt, and a management team that runs a simple business.
Give back with humility - Despite being one of the wealthiest individual, maintaining a modest lifestyle and an humble attitude towards his success with a strong commitment towards philanthropy.
The above eight principles strongly advocated by Buffett will be worth more to the investing world than the past or future returns of Berkshire Hathaway.
The Curious Cat would be summarizing the key lessons from each shareholder letter written by Warren Buffett every month in one of its sections.
Successful Brainstorming Done Remotely
Read the article here
The best method to generate ideas collaboratively is by brainstorming. With remote working environments, there are many approaches to solve complex problems.
Get a wide reach - Identify the roles and expertise you want, find people who fit that description, and ensure that the group you bring together is more diverse from various backgrounds bringing a range of diverse perspectives to solve the problem.
Take advantage of scheduling difficulties - Groupthink suggests that individuals think differently about a problem if they are alone, but tend to think alike together converging on a common solution. Collecting different ideas from small groups of people and then showcasing the collated list to everyone so that the group can discuss the most promising ideas and reach a consensus is an effective technique.
Get specific - The more distant you are from something in time, space or socially, the more abstractly you think about it. Abstraction can help in finding good analogies to provide insight but can also fail in providing specific solutions. Being specific about your ideas can help in providing clarity.
We hope these techniques do help in easing out the distance and help you get new and creative ideas to implement at work.
GPT -3: An Inflection Point in Artificial Intelligence
Read the article here
Generative Pretrained Transformer (GPT-3) is a program that accepts some text and then generates a continuation of the text. It mimics the style, tone and character of whatever text you feed it. GPT-3 was trained to build on a neural network that closely resembles the working of a human brain. A neural network learns by feeding it large amounts of data, GPT-3 was provided textual data available on the Internet. It has learnt the subtle rules that dictate how humans pick our words in different contexts. The GPT-3 is grabbing headlines because of its eerily similar resemblance to how a human actual responds to queries.
Language is the fabric of our society, our use and misuse of language dictates the rise and fall of human endeavour and currently a computer program has become really good at imitating the language.
The progress is huge but in the presence of a bizarre question, the response is equally bizarre. GPT-3 gives awkward garbage responses to extremely profound and intelligent questions like “What’s up?”
GPT- 3: The Hope from AI
Read the article here
Tyler Cowen writes that the unveiling of GPT-3 is a watershed moment in the history of Artificial Intelligence. The eventual uses of GPT-3 are hard to predict, but it is easy to see the potential. GPT-3 can converse at a conceptual level, translate language, answer email, perform (some) programming tasks, help with medical diagnoses and, perhaps someday, serve as a therapist. It can write poetry, dialogue and stories with a surprising degree of sophistication, and it is generally good at common sense — a typical failing for many automated response systems.
Some probable dangers involve the use of racial words can lead to unappetizing responses. A more advanced version of GPT-3 can be a powerful surveillance engine for written text and transcribed conversations. Considering all the positives and the drawbacks, in a grim year this is a positive development.
If you are employed in a job in any organization, it would help you to know the extent of disruption that an algorithm can cause in your line of work.
With a multitude of personal data and an algorithm that can accurately predict the way you speak, a combination of the two can help in automating many lower-order activities leaving humans to attempt more complicated stuff.
Job Interviews Don’t Work
Read the article here
Job interviews don’t work when the whole process is based on subjective feelings. They are in no way the most effective means of deciding who to hire because they maximize the role of bias and minimize the role of evaluating competency. Job interviews are representative of the fundamental attribution error. We view people’s behaviors as the visible outcome of innate characteristics, and we undervalue the impact of circumstances. Gut feelings are inaccurate and experience does not equal expertise in interviewing.
A way to make them effective involves building structured interviews. In a structured interview, everyone gets the same questions with the same wording, and the interviewer doesn’t improvise. Blinding the process by disguising key information that may lead to biased judgments. Blinding is another technique to reduce unconscious bias in interviews. Conducting competency related evaluations and work sample tests are effective supplements to make interviews effective.
When you understand the limitations and pitfalls of the job interview, you improve your chances of hiring the best possible person for your needs. And getting hiring right is paramount to the success of any organization.
Mental Model For The Week - Predictive Coding
It is a theory of brain function in which the brain is constantly generating and updating a mental model of the environment. It is used to generate predictions of sensed inputs. The human perception, like predictive text, replaces the unknown with the expected based on a database of past experiences.
There is no actual movement on the screen of the television, no actual spaces between spoken words, but it is the work of the human brain that encodes the gaps with its perceived input based on past experiences.
Afterthought
There are no whole truths;
All truths are half-truths.
It is trying to treat them as
Whole truths that plays the devil.
-Alfred North Whitehead, Dialogues (1953)
And that is a wrap for the week. We hope you enjoyed binge reading this edition. If you found this newsletter useful and worth your time, do share it with your friends. We would really appreciate some positive word of mouth.
We are eager to hear from our readers. Please reach out to us by sharing your feedback as a reply to this email. You can also send a DM @_vighneshnayak on Twitter with your suggestions to improve this newsletter.
Take care, stay safe and have a nice weekend. We shall see you next Saturday