Welcome back to
Thanks for voting and helping us shape these important conversations!
This week’s poll asked a clean, binary-feeling question: Are you good at spotting AI-generated content online - text, images, videos, deepfakes? The question hits different now than it would have two years ago. AI-generated text, images, and deepfakes have gotten so good that the initial usual tells are starting to disappear.
The results felt less like a poll and more like a confession booth.
This Week’s HackerNoon Poll Results
Are You Good at Spotting AI-generated Content Online? (Including Text, Images, Videos/Deepfakes)
Are you able to trust what you see online?
Most of Us Are Winging It
44% picked the honest answer: "I can sometimes spot it, but it's getting harder." These are the well-calibrated ones. Their instincts still work, but it’s undeniable that those instincts are being quietly outpaced by the constant development of chatbots and generative AI.
Only 34% said they're pretty good at detecting AI slops. Based on the majority of answers, that confidence is worth questioning and can be debated. It is true that as AI becomes a larger part of our daily lives, general familiarity can help users spot the hallmarks of machine-generated content. However, AI models are advancing rapidly. Soon, mere familiarity won't be enough to reliably catch AI-generated filler as we scroll through our feeds.
Then There's the 22%
Combine the 12% who said it's out of control with the 10% who no longer trust verified sources, and you've got nearly a quarter of respondents losing faith in the internet entirely. This is a clear sign of consumption exhaustion when the line between genuine and AI-generated content has been blurred exponentially.
Let’s focus on the claim that says “I don’t even trust verified sources now”. When users can't rely on historically credible publishers to filter out the noise, they stop trying to verify what is real and start assuming everything is fake.
The poll doesn't tell us how to fix it. But 66% of a technically literate audience admitting they're losing ground? That's a signal worth taking seriously.
Weigh in on the poll results
Around The Web - Kalshi’s Pick
Kalshi currently positions Anthropic as the clear favorite to win the 2026 AI coding race, giving them a dominant 47.8% probability. Meanwhile, OpenAI (22.2%) has gradually slipped into a dead heat with Google (22.8%) for second place, setting the stage for a tight battle over whose next major model release will close the gap.
Around The Web - Polymarket’s Pick
In contrast, Polymarket bettors predicting the short-term winner for "best AI model for coding on March 31" overwhelmingly favor OpenAI at 75%. The chart shows OpenAI breaking away from Anthropic (11%) in mid-February, while Google (8.4%) and DeepSeek (4.5%) trail significantly. This sharp divergence from the long-term Kalshi outlook suggests the market is pricing in an imminent, highly anticipated release from OpenAI that they expect will dominate the immediate conversation, even if Anthropic is still viewed as the safer bet for sustained leadership through 2026.
Join the Conversation
Vote in this week’s poll: Do you think 'software engineer' will still be a common job title in 10 years?
We’ll be back next week with more data, more debates, and more donut charts!