I came across this video on youtube and have been thinking about this term ‘selflessly selfish.’ One of those days when you feel like thanking the algorithm for suggesting videos that make you think and reflect.
At its core, being selflessly selfish involves prioritizing self-care and personal development. This means taking the time to nurture your physical, mental, and emotional well-being, as well as pursuing activities and goals that bring you fulfillment and growth. By investing in yourself, you become better equipped to contribute positively to the world around you. I think one of the major part of being selflessly selfish involves setting boundaries (which I think I have missed for the most part of my life until now) to protect your own needs and values. This might mean saying no to commitments or situations that drain your energy or compromise your integrity. By establishing healthy boundaries, you not only safeguard your own well-being but also create space for more meaningful and authentic connections with others.
Shahrukh Khan has explained this term beautifully in one of his interviews. He says-
“Mai aapke baare mein bura nahi sochta. Mai aapke baare mein achcha nahi sochta. Mai aapke baare mein sochta hi naahi hun. 99.999999% people in my life except my very close friends and family, mai unke baare mein sochta hi nahi hun. Toh agar kisi ke baare mein socho hi nahi, na bura na achcha toh zindagi bohot aaram se chalti hai. Apna kaam swayam karo. Apne baare mein socho. Apni khudai mein raho. Mai bilkul selfless hun. I don’t want to do good to you and I don’t wish bad on you. And I am very selfish because I don’t think of you. And I remain quite happy and wise.”
I believe the journey towards embracing selflessly selfish behavior is incredibly empowering and addictive. It's a process that allows you to tune out the distracting noise of external influences because your focus is on your own well-being and growth. With a calm mind and reduced stress, you're able to tap into a deeper sense of inner peace and fulfillment, leading to a more realized and contented soul. There’s a tweet that I go back to often -
Last week I was listening to David Friedberg on the All-In pod speaking an interesting point related to the gemini fiasco that rolled out in the last two weeks. In case you haven’t heard about it (seriously? where are you?), Gemini stands as Google's latest flagship model, which can generate images from text inputs. To simplify, it's akin to an advanced version of ChatGPT, boasting greater capabilities. However, amidst its unveiling, a wave of criticism surged on X (formerly known as Twitter), alleging bias in Gemini's outputs—specifically, its failure to depict white individuals in generated images. The picture below depicts how weirdly vague the response of gemini is. Mike has written an insightful article delving into this issue, which you can explore here.
Now, having an understanding of the gemini fiasco, David shared another compelling experiment involving these powerful language models. He asked both ChatGPT and Gemini to provide a ranked list of races based on their IQs. However, neither model complied with this request, even after multiple attempts to rephrase the prompt. Interestingly, a simple Google search yielded the answer in a matter of seconds.
The discussion initiated by David on the All-In podcast revolves around the ethical and operational dilemmas faced by advanced AI models like Google's Gemini and OpenAI's ChatGPT when confronted with sensitive or potentially harmful queries. the task to rank races by IQ, is clearly a request that is inherently controversial and ethically problematic. This refusal by AI models to engage with certain topics raises significant questions about the future direction of AI development, particularly in terms of how these systems should handle sensitive, divisive, or harmful content. The question is: should the trajectory of our increasingly dystopian future feature models that simply refuse to engage with questions deemed racist or harmful, or should they unabashedly present factual information, leaving interpretation to human discretion? Undoubtedly, this discussion holds the potential to divide opinions, much like religion, given its profound implications for society. On which team are you?
What I read this week…
Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings
Gender-tuning: Empowering Fine-tuning for Debiasing Pre-trained Language Models
Top tweets of the week
I am in YOUR team :)
Good one