It feels like a decade has passed, but we're really only looking at the timeline from May 2023 to now. Back then, everyone was freaking out because GPT-4 was still shiny and new, and people thought their jobs would vanish by Tuesday. We were all staring at prompts, wondering if a chatbot could actually replace a junior developer or a copywriter. Well, it didn't happen exactly like the doomsday prophets said, but man, the world definitely looks different.
The Shift from Chatbots to Agents
In mid-2023, the conversation was mostly about "chatting." You'd ask a bot a question, it would give you a slightly hallucinated answer, and you'd move on. But then things started getting weirdly functional. Companies stopped just releasing models and started building ecosystems. Open AI’s DevDay in late 2023 was a massive turning point because it introduced GPTs—basically custom versions of the tech that could actually do stuff.
Think about it. We went from "Write me a poem about a toaster" to "Here is my entire tax history, please find the deductions I missed and draft an email to my accountant."
The technical term is "agentic workflows." Andrew Ng, a guy who actually knows what he’s talking about in the AI space, has been shouting from the rooftops about this. He basically argues that instead of just asking an AI to do something once, we’re now letting it iterate. It writes, checks its own work, fixes the bugs, and then gives you the final product. This change, which really gained steam throughout 2024 and into 2025, is why the tech feels more "useful" and less like a parlor trick today.
The Hardware Reality Check
Remember the Humane AI Pin? Or the Rabbit R1? Those were supposed to be the "iPhone moments" of 2024.
They weren't.
They were basically expensive paperweights that proved one thing: we still like screens. Most people realized that having an AI whisper in your ear or project a grainy laser onto your hand is kinda annoying compared to just using a phone. However, it wasn't all a bust. Meta’s Ray-Ban glasses actually caught on because they looked like regular glasses but could tell you what you were looking at. It's a subtle difference, but it matters.
💡 You might also like: The H.L. Hunley Civil War Submarine: What Really Happened to the Crew
The Lawsuits and the "Data Wall"
You can't talk about the period from May 2023 to now without mentioning the massive legal cage match happening behind the scenes. The New York Times sued OpenAI. Artists sued Midjourney. Authors like Sarah Silverman got involved.
Basically, the "Wild West" era ended.
These lawsuits are forcing a massive shift in how these models are trained. We're seeing a move toward "synthetic data." Since the internet is basically full of AI-generated junk now, the models are starting to train on data created by other models. It's a bit like a copy of a copy of a VHS tape—eventually, the quality starts to get grainy. This is a real problem that researchers at places like Oxford and Cambridge have called "model collapse." If we don't fix it, the AI of 2026 might actually be dumber than the AI of 2023.
Big Tech’s Identity Crisis
Google panicked. Let's be honest. When Gemini (formerly Bard) first rolled out, it was... fine. But then it had some high-profile hiccups with image generation that turned into a PR nightmare. Apple, on the other hand, played the long game. They waited until mid-2024 to announce "Apple Intelligence," focusing on privacy rather than raw power.
It was a smart move.
Most people don't care about the underlying parameters or the "FLOPs" of a model. They just want their phone to remember their mom's flight number and tell them if they have time to grab coffee before picking her up. This shift toward "Local AI"—stuff that runs on your actual device rather than the cloud—is the biggest technical trend of the last eighteen months.
📖 Related: The Facebook User Privacy Settlement Official Site: What’s Actually Happening with Your Payout
Why the Economy Didn't Collapse (Yet)
There was a lot of talk in May 2023 about mass unemployment. While some sectors, particularly entry-level coding and basic content writing, have seen a massive squeeze, we haven't seen the "Great Displacement" yet.
Why? Because implementing this stuff in a big company is a total nightmare.
I’ve talked to CTOs at Fortune 500 companies who say they spend 10% of their time on the AI itself and 90% of their time worrying about data security and compliance. You can't just plug ChatGPT into a bank's internal database and hope for the best. Regulations like the EU AI Act, which became a real thing during this window, have put guardrails on how companies can use this tech. It turns out, "moving fast and breaking things" is a bad strategy when "things" includes people's private medical records or financial data.
The Energy Problem Nobody Saw Coming
Here’s something people often miss: AI is thirsty. Not for water (though it uses a lot of that for cooling), but for electricity. Since May 2023, the sheer amount of power needed to run these massive data centers has become a geopolitical issue. Microsoft is literally trying to restart a nuclear reactor at Three Mile Island to power their AI ambitions.
Think about that for a second. We’re reviving Cold War-era nuclear infrastructure just so we can generate more high-res cat pictures and automated emails. It's wild.
What Actually Works Now
If you're looking for what has actually changed in the day-to-day for a regular person since May 2023 to now, it’s the small stuff.
👉 See also: Smart TV TCL 55: What Most People Get Wrong
- Video Translation: You can now film a video in English, and an AI can change your voice and your lip movements to make it look like you're speaking perfect Spanish or Mandarin.
- Coding for Non-Coders: Tools like Cursor or Replit have made it so someone who barely knows HTML can build a functional app in a weekend.
- Personalized Learning: Khan Academy’s Khanmigo showed that AI can actually be a decent tutor, asking students questions rather than just giving them the answers.
It’s not a revolution; it’s an evolution. The tech has become "invisible." It's in your email autocomplete, your photo editing tools, and your spreadsheet formulas.
The Mid-2023 Hype vs. 2026 Reality
In 2023, we were promised AGI (Artificial General Intelligence)—a machine that can do anything a human can. We aren't there. Not even close. What we have instead is "Specialized Competence."
The models are great at patterns. They are terrible at logic. If you ask an AI to solve a complex math word problem it hasn't seen before, it still trips over its own feet. But if you ask it to summarize a 50-page PDF, it’s a godsend. We've learned to stop treating it like a digital god and start treating it like a very fast, slightly eccentric intern.
Actionable Steps for the New Era
If you've been sitting on the sidelines since May 2023, it's time to stop. The "wait and see" period is over. Here is how you actually handle this shift without losing your mind:
Focus on "The Last Mile"
AI can do the first 80% of almost any digital task. Your value now lies in that final 20%—the fact-checking, the nuance, the "human" touch that keeps it from sounding like a corporate brochure. If you just copy-paste from a prompt, you're replaceable. If you use the prompt as a foundation and build something unique on top of it, you're indispensable.
Audit Your Workflow
Look at the tasks you do every day. Anything that involves "summarizing," "reformatting," or "initial drafting" should probably be offloaded. Use the time you save to do the things AI can't do: build relationships, think strategically, and solve problems that don't have a pre-existing pattern on the internet.
Get Comfortable with Multimodality
Stop just typing text. Start using voice. Start using image inputs. The newest models see and hear the world. If you're a mechanic, take a photo of a weird engine part and ask the AI to identify it. If you're a teacher, record your lecture and have the AI generate a quiz based on the specific points you made.
The world changed a lot between May 2023 to now, but it didn't end. We just got a new set of tools. They’re messy, they’re power-hungry, and they’re occasionally full of it, but they’re here to stay. The people who win in the next few years won't be the ones who "mastered" AI, because the tech changes too fast for that. It'll be the people who remained curious and adaptable enough to keep up with the mess.