Banner

30 Things I've Learned About AI:In 8 years of studying and writing about it


Don't just trust headlines about AI tools—try them out yourself.
After any AI bubble bursts, survivors will create a new tech and cultural landscape.
AI itself won’t replace your job; someone who uses AI will—possibly even you, if you don’t adapt.
Different stakeholders prioritize different things: shareholders, votes, views, or citations. Find balanced sources.
Generative AI’s final phase isn’t valuable for what it creates, but for what it leaves out.
AI is not just futuristic—it’s also part of the past and present; once integrated into daily life, we stop calling it “AI.”
AI needs regulation, but innovation shouldn't suffer for it.
Some people predicted today’s AI advances a decade ago. Following them can give you a glimpse into the future.
AI is both marketing and not marketing—it’s nuanced.
Extreme hype and anti-hype often come from similar sources.
During conflicts, pay attention to quiet voices; loud ones are often taking sides.
Understanding neural network math offers unique insights.
AI’s trajectory may surprise you in months—keep reflecting on it.
AI won’t always be this cheap; this is an unusual period.
Testing AI tools is low-cost with high potential benefit. Only those deeply invested in opposing AI avoid it.
GPT-5’s impact will be unexpected—technology is constantly evolving.
You’re not the target of AI hype; it’s investors that companies need to keep the market alive.
Generative AI is a fraction of AI; much of it is predictive, like spam filters and forecasting.
In a decade, most writers will likely use AI in their work.
AI has become like politics—people often hold firm beliefs regardless of new evidence.
Knowing about AI doesn’t guarantee instant wealth.
AI x-risk believers genuinely worry about potential dangers; tech-optimists often have personal interests at heart.
Sam Altman seeks a lasting legacy more than money, for better or worse.
AI data can be skewed to support any stance—beware of single studies.
China may lead AI if the U.S. doesn’t adopt a more unified approach.
Most people are indifferent or unaware of AI, while a few are either passionate or hostile about it.
People enjoy human elements in content; purely AI-made content often needs deception to succeed.
To excel in AI, you need curiosity, practice, and a willingness to update your understanding.
Your perception of AI will evolve over time—and that’s perfectly normal.