A growing crisis of trust plagues the internet, extending beyond mere AI ‘slop’ on social media. Artificial intelligence is increasingly generating text and images found in job applications, product evaluations, and even insurance claims, causing widespread confusion for both platforms and users struggling to distinguish authentic content.
Over the last few years, several startups, like Pangram, have emerged to establish a crucial ‘trust layer’ for the internet. Pangram recently secured $9 million to advance its AI detection technology and formed a partnership with Substack. Through this collaboration, Substack now employs Pangram’s tools to inform readers when their preferred authors utilize AI in their newsletters. Furthermore, Pangram has also introduced a new tool for detecting AI-generated images.
This installment of TechCrunch’s Equity podcast features Pangram co-founder and CEO Max Spero, who joins Rebecca Bellan. They delve into the potential of AI detection technologies and discuss the critical distinction between content that is merely AI-assisted and that which is fully AI-generated.
The complete episode offers further insights into:
- Spero’s perspective on why the internet might be “dangerously close” to embodying the dead internet theory in the near future.
- The challenges and greater utility in quantifying AI’s contribution to content, rather than simply categorizing it as AI or human.
- The significant implications of inaccurate AI detection, particularly regarding false positives with sensitive imagery.
- Spero’s view that lower-tier writing positions might vanish permanently, while authentic, high-quality human writing could gain increased value.
Access Equity by subscribing on YouTube, Apple Podcasts, Overcast, Spotify, and other major podcast platforms. You can also follow Equity on X and Threads via @EquityPod.