“The more entangled you are in that identity, the harder it is step back because we need distance to see things. The power of hindsight is particularly useful and strong because we can use it to look at the past, but when you're in the thick of things, it's very, very difficult to step away.”
Listen now:
Madalina Buzdugan has spent her entire career in tech startups and scale-ups, working at the intersection of sustainability and AI strategy. Her squiggly career started in communications and PR, moved through event management and global mobility work in Norway, then into expansion strategies for a travel startup where she spent 80% of her time on the road. After taking time off, she transitioned to the EV space doing sustainability work focused on carbon footprints and life cycle assessments. Following another career break, she now explores what sustainable AI means and how to apply AI responsibly while maintaining her foundation in corporate sustainability.
📚 What You'll Learn
Why work identity has become dangerously entangled with self-identity for millennials, making career breaks feel taboo and risky
How generative AI can't have positive climate impact due to resource consumption, but predictive AI and machine learning are creating impressive climate solutions
Why batching your prompts and being concise can dramatically reduce AI's energy consumption and environmental impact
✍️ Some Takeaways
The taboo around taking time off stems from work identity being entangled with self-identity, especially for millennials who were told "work hard and everything will work out." Layered with family sacrifices and being first-generation to get higher education. Tech's "we're a family" culture with pizza Fridays intensifies the overlap. When you've given so much to reach seniority, stepping away feels like enormous risk because you don't know who you are outside that status. The fast rhythm makes it worse - technology shifts constantly, and a week away means thousands of emails.
Shift how you react to career breaks by responding with "Congratulations" instead of concern, and seek out people who've taken time off to understand their experience. Look for people in your network who've stepped away and invite them for coffee. Without her six-month break, Madalina wouldn't have had the mental space to study AI deeply or develop the expertise to now speak on stages about sustainable AI.
Generative AI cannot contribute positively to climate goals, but predictive AI and machine learning are creating impressive climate solutions that deserve more capital. Large language models use enormous energy and water with no positive climate impact. But predictive AI enabled Indian farmers to use satellite data for optimal irrigation timing, helping yields and soil health. In the Amazon, AI on brick phones listens for illegal logging and alerts rangers. These targeted applications solve real problems without generative AI's resource consumption - this is where capital should flow.
You can reduce AI's environmental impact immediately by batching prompts into single requests and being as concise as possible to eliminate guesswork. Every prompt revs up a powerful engine. Asking for a green slide, then larger font, then orange background fires that engine three times. Take five minutes to compile requests into one prompt - you fire up once instead of ten times. Similarly, avoid vague rambling. Be specific and concise because the less the AI has to guess, the less computational power and energy it uses.
Links
Where to find Madalina Buzdugan
Where to find Milly
Website: http://www.millytamati.com/
Generalist World Resources
🙏 Special thanks to our podcast producer James McKinven! (get in touch for all your podcast needs, he’s really great!)
