Chapter 7: Ways AI Has Changed Since We Started HonestAI
7. 5 Ways AI Has Changed Since We Started HonestAI
When we launched HonestAi , artificial intelligence was riding a wave of optimism, anxiety, and a fair amount of science fiction. The promise was enormous, the risks were unclear, and the future felt just close enough to imagine but far enough to debate.
Years later, the world has changed and so has AI.
From boardrooms to classrooms, smartphones to operating rooms, artificial intelligence is no longer emerging. It has arrived, and with it, so have profound questions about power, ethics, creativity, and identity.
In this special anniversary retrospective, we reflect on how AI has evolved—not just in terms of technology, but in how society thinks about it, talks about it, and governs it. Here are five major shifts we’ve witnessed since the first issue of HonestAI hit the stands.
Table of Contents
7.1. The Hype vs. Reality Curve: Growing Up Fast
Few years ago, the headlines were filled with promises of sentient machines and overnight revolutions. The reality that followed has been more nuanced and more human.
Yes, generative AI experienced explosive growth. Breakthroughs in large language models like GPT-4o, Claude, and Gemini have redefined what’s possible and rewritten the digital playbook. But with those advances came challenges: misinformation loops, hallucinations, massive compute demands, and complex ethical dilemmas. The initial hype cycle is giving way to something more mature, a period of critical engagement and reflection.
Then (2022): “AI will take your job.”
Now (2025): “How can AI augment your expertise—without compromising human agency?”
We’ve come to understand that technological maturity doesn’t eliminate uncertainty. Instead, it sharpens the questions we ask and deepens the conversations we need to have.
7.2. From Vision to Regulation: AI Policy Becomes Real
In 2022, conversations about AI policy were largely aspirational—focused on broad visions of ethical principles, innovation, and responsible development. But today, in 2025, policy has caught up with technology.
AI is no longer just a technical challenge; it’s a matter of law, governance, and public accountability.
The EU AI Act—passed in 2024—became the world’s first comprehensive AI regulatory framework, classifying AI systems by risk and imposing transparency and safety requirements.
The U.S. AI Executive Order in 2023 triggered waves of policy development around fairness, cybersecurity, and public sector AI use.
China has enforced strict algorithm regulations, including real-name algorithm registries and content control.
Regulators now shape the roadmap, not just respond to it. And public trust in AI increasingly hinges not just on what it can do—but who governs it, and how.
Fun Fact: As of early 2025, more than 35 countries have enacted some form of national AI strategy or legislative action.
7.3. When AI Fades Into the Background
Look around, and you might not see it, but AI is everywhere. It hums quietly beneath the surface of your digital life, making decisions for you before you even notice they need to making:
70% of the videos you see on TikTok are ranked and served by AI algorithms.
Even city traffic systems in smart urban areas increasingly rely on AI to manage flows and reduce congestion.
What once felt like futuristic demos—robotic assistants, language models, or AI-powered chess engines—have evolved into something subtler and far more pervasive. AI isn’t just a feature anymore. It’s the operating logic of the modern world.
This ambient AI, unseen but constantly present, represents both a triumph and a challenge.
We’ve reached a moment where children interact with AI—voice assistants, recommendation engines, smart toys, before they even learn cursive handwriting.
This quiet normalization of AI into daily routines is not just a technical milestone.
It’s a cultural shift as one that will shape how future generations understand knowledge, autonomy, and trust.
In this new era, AI no longer feels “cutting-edge.” It feels like infrastructure.
And that makes it more powerful and more important to scrutinize—than ever before.
7.4. The Great Explainability Push
Five years ago, most AI systems were “black boxes.” They made decisions, but no one really knew how or why. That used to be normal. Today, things are different. Explainable AI (XAI) is now a key priority in both research and real-world products.
In industries like healthcare, finance, and criminal justice, people need to understand how AI systems work—especially when lives, money, or legal outcomes are on the line. That’s why interpretable models are now preferred in these fields.
To meet this need, companies are doing more to explain their AI systems:
They’re using model cards and data sheets to show how models were trained and what they’re meant to do.
Tools like OpenAI’s system message transparency and Anthropic’s Constitutional AI give users a better idea of how decisions are made.
And here’s a big number:
FUNFACT :- 73% of enterprise clients in regulated industries now require AI systems to include an explanation layer .
So what changed?
Explainability is no longer just for researchers or ethics experts. It’s now a business must-have and a legal requirement in many places. If we want to trust AI, we need to know not just what it does—but how and why it does it.
The past five years have shown us that AI is neither savior nor villain. It is a mirror of human intent and a magnifier of our values, for better or worse.
As we look ahead, our mission remains the same: to crowd source insight, challenge assumptions, and make AI more honest—together.
Contributor:
Nishkam Batta
Editor-in-Chief – HonestAI Magazine
AI consultant – GrayCyan AI Solutions
Nish specializes in helping mid-size American and Canadian companies assess AI gaps and build AI strategies to help accelerate AI adoption. He also helps developing custom AI solutions and models at GrayCyan. Nish runs a program for founders to validate their App ideas and go from concept to buzz-worthy launches with traction, reach, and ROI.