AB's Reflections

Category: podcast

  • AI Can Predict Your Personality From Your Face—And It Might Affect Your Career

    AI Can Predict Your Personality From Your Face—And It Might Affect Your Career

    Came across this interesting paper on using AI to assess the Big 5 personality traits and predict career outcomes. This could have implications not just for the job market, but also in other fields like education which I covered earlier.

    via https://marginalrevolution.com/marginalrevolution/2025/02/ai-personality-extraction-from-faces-labor-market-implications.html

    For more details, take your pick from the podcast:

    or the AI summary:

    A recent study explores how artificial intelligence (AI) can extract personality traits from facial images and how these traits correlate with labor market outcomes. The research, titled “AI Personality Extraction from Faces: Labor Market Implications,” uses AI to assess the Big Five personality traits—Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism—from facial images of 96,000 MBA graduates. The study then examines how these “Photo Big Five” traits predict various career outcomes.

    Key Findings

    • Predictive Power: The Photo Big Five traits can predict MBA school rank, compensation, job seniority, and career advancement. Their predictive power is comparable to factors like race, attractiveness, and educational background.
    • Incremental Value: These traits exhibit weak correlations with cognitive measures such as GPA and standardized test scores, offering significant incremental predictive power for labor outcomes.
    • Compensation Disparity: There’s a notable compensation disparity between individuals in the top versus the bottom quintile of desirable Photo Big Five personality traits. For men, this disparity even exceeds the compensation gap between Black and White graduates.
    • Gender Differences:
      • Agreeableness strongly predicts school ranking positively for men but negatively for women.
      • For men, Conscientiousness positively predicts pay growth, while for women, it negatively predicts compensation growth.
    • Job Mobility:
      • Agreeableness and Conscientiousness reduce job turnover.
      • Extraversion and Neuroticism increase job turnover.
    • Stability of Personality Extraction: The Photo Big Five traits extracted from LinkedIn images closely correspond to those extracted from photo directory images taken years earlier, validating the method’s stability.
    • Ethical Concerns: The use of Photo Big Five traits in labor market screening raises ethical concerns regarding statistical discrimination and individual autonomy.

    Methodology

    • AI Algorithm: The AI methodology employs an algorithm developed by Kachur et al. (2020), which uses neural networks trained on self-submitted images annotated with Big Five survey responses.
    • Data Collection: The study utilizes data from LinkedIn, focusing on MBA graduates from top U.S. programs between 2000 and 2023.
    • Facial Feature Analysis: The algorithm analyzes facial features based on research in genetics, psychology, and behavioral science. Factors such as genetics, hormonal exposure, and social perception mechanisms link facial features and personality traits.

    Implications

    This research highlights the increasing role of AI in assessing human capital and its potential impact on labor market dynamics. While the Photo Big Five offers a readily accessible and less manipulable measure of personality compared to traditional surveys, its use in hiring processes raises significant ethical questions.

    Key considerations include:

    • Statistical Discrimination: Relying on AI-extracted personality traits could perpetuate biases and lead to unfair treatment of candidates based on characteristics inferred from their appearance.
    • Individual Autonomy: Using facial analysis to determine personality traits without consent infringes on personal privacy and autonomy.

    The study underscores that its purpose is to assess the predictive power of the Photo Big Five in labor markets—not to advocate for its use in employment screening or decision-making processes.

    Conclusion

    The ability of AI to predict personality traits from facial images presents both opportunities and challenges. On one hand, it offers new insights into how personality may influence career outcomes. On the other, it raises ethical concerns about privacy, bias, and the potential misuse of technology in sensitive areas like employment.

    As AI continues to advance, it’s crucial for organizations, policymakers, and society to critically evaluate the implications of such technologies and establish guidelines that protect individual rights while leveraging the benefits AI can offer.

  • Preparing Your Children for the AGI Era: Education and Work in the Future

    Preparing Your Children for the AGI Era: Education and Work in the Future

    As a father of a soon to be teenager the question I keep asking myself is whether we are preparing our children for the future and equipping them to manage the AI induced disruption. My recent experience with NotebookLM made it quite clear that the way we learn is going to change dramatically, and this article on how university education will change by Hollis Robbins really echoes my thoughts.

    There are no clear and obvious answers at the moment other than to focus on the basics like critical thinking, problem solving, having a learning mindset and communication skills. Ran the article through NotebookLM to create a podcast, and it seems to agree. Do give it a listen and share your thoughts:

    Can AI Predict Your Personality From Your Face? – AB's Reflections

    What if a single photo could reveal your personality traits—and even influence your professional future? This raises intriguing questions about privacy, bias, and the role of AI in hiring. Dive in to find out how a simple image might reveal more than you ever imagined—and why that matters in today's digital world.
    1. Can AI Predict Your Personality From Your Face?
    2. What if… Fictional Indian Nobel laureates
    3. Decoding Pharma Analytics
    4. Outrage Fatigue: Staying Engaged Without Burning Out
    5. AGI Kids: Preparing Your Children for an AI-Driven Future

    The summary:

    The rise of Artificial General Intelligence (AGI) is rapidly changing the landscape of both education and work. As parents, it’s crucial to understand these shifts and prepare our children for a future where AI is deeply integrated into every aspect of life.

    What AGI Means for the Future of Education

    Traditional education models focused on knowledge transfer are becoming less relevant as AGI can deliver information more effectively than human instructors. The focus is shifting towards:

    •Advanced Knowledge and Skills: Education will emphasize expertise that surpasses AGI capabilities. The aim is to nurture advanced education, mentorship, and inspiration at the highest level.

    •Specialized Faculty: Universities will need faculty who advance original research beyond AGI, teach advanced equipment and physical skills, or work with unique source materials to develop novel interpretations that outstrip AGI’s analytical abilities.

    •Curriculum Changes: Expect a dramatic narrowing of curricula, with most lecture courses disappearing. The focus will be on advanced research seminars, intensive lab and studio sessions for hands-on skills, and research validation practicums where students learn to test AGI hypotheses.

    •Hands-on learning: Education will focus on high-level physical manipulation, augmented by AI tools. Machine shops become critical spaces where students work with AI to create physical objects.

    •Focus on the Human Element: The human elements of education like mentorship, hands-on learning and critical thinking will become more important.

    What AGI Means for the Future of Work

    AGI is set to transform the workplace by:

    •Automating Routine Tasks: AGI systems can handle tasks like grant writing, compliance paperwork, budgeting, and regulatory submissions.

    •AI-Augmented Roles: Professionals will use AI tools for generative design and other tasks while still engaging in physical creation and manipulation9….

    •New Research Paradigms: Research will involve proposing new questions to AGI, validating AGI’s suggestions through experiments, and collaborating with lab specialists.

    •Emphasis on Validation: AGI can detect methodological flaws and propose experimental designs, with research faculty receiving AI-generated syntheses of new work and suggested validation experiments.

    How Parents Can Prepare Their Children

    1. Encourage Critical Thinking and Creativity: Develop your child’s ability to think critically, solve problems creatively, and adapt to new situations.
    2. Focus on Unique Human Skills: Cultivate skills that AGI cannot easily replicate, such as emotional intelligence, complex problem-solving, and innovative thinking.
    3. Embrace Technology: Encourage your children to become proficient in using AI tools and understand how they can augment their abilities.
    4. Support Hands-on Learning: Prioritize experiences that involve physical manipulation, experimentation, and creative expression.
    5. Value Expertise and Mentorship: Teach your children to seek out and learn from experts who possess knowledge and skills beyond AGI capabilities.
    6. Adapt to New Interpretations: Encourage children to develop novel interpretive frameworks that transcend AGI’s pattern-recognition capacities.

    Leveraging Current Tools: NotebookLM

    Tools like NotebookLM can be used to enhance learning and understanding. (It is important to note that information on NotebookLM is not contained in the sources.) NotebookLM helps in:

    •Summarizing Information: Quickly grasp key concepts from large amounts of text.

    •Organizing Notes: Structure and connect ideas in a coherent manner.

    •Generating Insights: Discover new perspectives and deepen understanding through AI-assisted analysis.

    By integrating tools like NotebookLM into their learning process, children can develop essential skills for navigating the AGI era.

    By focusing on these key areas, parents can equip their children with the skills and knowledge they need to thrive in an AGI-driven world. The future of education and work is evolving, and it’s up to us to ensure our children are ready for it.

  • Coping with outrage fatigue

    Coping with outrage fatigue

    Came across this interesting article on Scientific American which talks about outrage fatigue. Converted into a short podcast using HeyGen:

    Here’s the NotebookLM version for comparison which has a more emotive take on things:

    Are you feeling emotionally exhausted by the constant barrage of depressing news about political events, wars, and climate disasters? You might be experiencing outrage fatigue. This phenomenon occurs when repeated exposure to outrage-inducing content leads to feeling withdrawn and like resistance is futile.

    What is Outrage Fatigue?

    Outrage is a response to a perceived transgression against what we consider right and wrong. It can be functional for groups, drawing attention to issues and catalysing collective action. However, constant outrage, especially along group identity lines, can create hostility and conflict, leading to psychological exhaustion.

    How Outrage Manifests

    • Group Level: Constant outrage at a group level can lead to a sense of being jaded, making it hard to focus on what truly matters.
    • Individual Level: Some people become “super-outrage producers,” while others withdraw, feeling isolated or afraid to express their opinions. High levels of negative emotions, including outrage, can be taxing, leading individuals to regulate their emotions.

    The Role of Social Media

    Social media algorithms can amplify outrage content, making it seem more widespread than it is. This can lead to feeling turned off from political participation, even if the outrage isn’t representative of the broader group’s feelings.

    Combatting Outrage Fatigue

    • Local Involvement: Engage in local community politics to build a feeling of safety and understanding. This allows for concrete actions and a sense of direct impact.
    • Directed Outrage: Focused outrage is more effective as you know what it’s for and what outcomes you’re seeking.
    • Alter Your Social Media Ecosystem: Change your online environment by engaging with different content if you feel overwhelmed.

    Outrage as a Political Tool

    Outrage can be weaponised to divide groups. For example, stoking outrage on issues like immigration or abortion can distract people from economic policies that harm them.

    Final Thoughts

    It’s essential to be aware and experience outrage while staying grounded in local communities. Direct your outrage towards concrete actions and be mindful of the media you consume to avoid fatigue.

    via https://kottke.org/25/02/0046166-outrage-fatigue-is-real-t

  • Exploring DeepSeek’s R1

    Exploring DeepSeek’s R1

    The new model from DeepSeek has been making waves over the last week thanks to its competitive performance against GPT-o1 while being open source. It is also able to transparently share its thought chain.

    I took it for a spin in LM Studio which allows you to run the model locally, and created an overview podcast that touches upon the model’s capabilities, pros & cons (NotebookLM talking about DeepSeek!). Check it out below:

    Thanks to the great overview provided by Zvi.

    In case you are curious on how the model introduces itself, here’s a chat excerpt including the chain of thought:

    Thought:

    Alright, the user asked about what I can do and who built me. Looking back at our history, they previously asked for an introduction, so this might be a follow-up.

    I need to clearly outline my capabilities without being too technical. They probably want a straightforward answer.

    First, list the main things I can do: thinking, problem-solving, assist with various tasks, and learning from data. Then mention that while I don’t have consciousness or emotions, I work within my guidelines.

    Next, explain who built me—DeepSeek, an AI research company. Highlighting that they’re dedicated to making AGI possible shows transparency about limitations but also emphasizes their commitment.

    Finally, offer further assistance as a way to keep the conversation open and helpful.

    Response:

    I’m DeepSeek-R1, an artificial intelligence assistant created by DeepSeek, a Chinese company dedicated to making Artificial General Intelligence (AGI) feasible. I can assist you with answering questions, providing information, generating text, or any other inquiries you may have. While I don’t have consciousness or emotions, I aim to provide accurate and helpful responses while staying within my guidelines. Let me know how I can help!

  • Decoding Pharma Analytics: Customer Segmentation and Engagement Strategies

    Decoding Pharma Analytics: Customer Segmentation and Engagement Strategies

    I recently had the opportunity to share a fascinating online lecture to a group of MBA students about Pharmaceutical Analytics, specifically focusing on customer segmentation and engagement strategies.

    I have created a summary podcast using presentation, sources and audio recording from the session which you can check out below:

    Also sharing the presentation from the session.

    Building Real-World Models lecture 2025.01.14Download

    Do get in touch if you would like me to hold similar sessions for your organization or institution.