Saved by a question
The tech investment landscape is changing dramatically, and recently, I had a meeting that highlighted this paradigm shift. Three tech investors and I dove into the heart of current trends, yet we found ourselves tangled in the weeds of technical experience and capability. I don’t think my responses enthused them. E.g. processes to establish value propositions or experience in Fintech. However, an unexpected question work breathed life into our meeting.
As the call and energy were dying, one investor asked, “Can I ask a question”.
We all paused, waiting. And then he asked, “What do you think of remote work?”
In this current climate, we, including his colleagues, assumed he would ask about my views on artificial intelligence and ChatGPT.
I was even prepping my response.
Relieved to be speaking about something interesting. I couldn’t stop talking.
Redefining the Workspace: Hybrid vs Remote
We’ve all heard the phrase ‘Remote work is the future,’ but how true is this in practice? The founders of Basecamp have posted that a hybrid model, which combines office and remote work, potentially stifles the benefits of a fully remote setting. I believe that complete immersion, whether 100% office or 100% remote, yields the best results.
Moreover, I shared my experience with a company that was not just remote but fully nomadic. This business model allowed their team to work asynchronously across different time zones, instilling trust and job satisfaction. This asynchronous, nomadic approach significantly impacted their rapid business growth.
AI and the 12 Cent Engineer
The highlight of our conversation revolved around the emergence of AI in coding, specifically the “12 Cent Engineer”. This concept, presented by Fixie.ai’s CEO Matt Welsh, argues that AI can potentially replace a Silicon Valley dev, costing only 12 cents a day. Welsh predicts that AI could translate high-level natural language into quality code in three years, leading to up to 100 million times in cost savings and efficiency improvements.
Although some may dismiss AI capabilities as mere hallucinations producing subpar code, my experience with tools like GitHub’s Copilot suggests otherwise. It often feels like Copilot is reading my mind, guiding me to better solutions.
Reconsidering Early-Stage Investment
With the rise of AI in coding, the tech investment landscape will inevitably transform. Investors might soon find themselves backing teams of visionaries, capable of churning out hundreds of new ideas supported by AI tools, rather than the conventional mix of a visionary leading a team of ‘fact-finding’ software developers. It’s a shift that asks us to reconsider what an ideal early-stage team looks like.
AI now builds 49% of New Code on GitHub
According to a few Podcasts, I’ve listened to, 49% of new code on GitHub is generated by AI. This is remarkable, with GitHub Copilot now responsible for an average of 46% of a developer’s code across all programming languages and up to 61% in Java.
GitHub continues to upgrade its AI-powered coding assistant, with key improvements such as a better AI Codex model for superior code synthesis and an AI-based vulnerability prevention system that identifies and blocks insecure coding patterns. These enhancements provide an end-to-end experience for developers to secure their code directly from their text editors.
The rate of tech evolution is dizzying, and AI’s role in coding is no exception. As one of the investors aptly responded after our discussion, ‘Wow.’
It certainly changed a pretty standard sales meeting into an exciting and engaging debate.
A few related posts: