past-years-ai-advances
Technical Thoughts

The Past Two Years of AI: Why We’re Undervaluing an Incredible Leap

Artificial Intelligence has come a long way in just two years. From remarkably realistic chatbots to AI-powered research assistants, the pace of innovation has been nothing short of astounding. Yet despite these leaps, many people still underestimate how far AI has progressed—and how deeply it’s shaping our lives.

This post takes a closer look at the most notable developments in AI over the past couple of years and why we might not be giving them the credit they deserve. We’ll also touch on Levlex, a next-generation AI platform that ties many of these advancements into one powerful system.

1. Chatbots Moved From Gimmick to Genuine Assistants

Just two years ago, most chatbots were stuck answering basic FAQs or following heavily scripted flows. Fast-forward to today, and we have large language models capable of:

They’re more than mere “chat” companions; they’re now capable of advanced content creation, strategic analysis, and brainstorming. And while the general public still tends to see them as novelty toys, entire industries—from legal research to content marketing—are being transformed by these AI assistants.

Why We’re Undervaluing It
There’s a natural human tendency to normalize the incredible—after seeing AI do something impressive a few times, we label it “cool” and then quickly move on. But next time your AI helper drafts a coherent blog post or references a key fact from thousands of words ago, remember: we’re witnessing capabilities that seemed like distant science fiction just two years back.

2. Natural Language Processing Broke Its Own Records

Over these past two years, models like GPT-3.5, GPT-4, and various open-source LLMs (like LLaMA) have smashed records in tasks once considered near-impossible:

With these breakthroughs, AI is no longer limited to short, simple interactions—it can fully participate in extended discussions, keep track of complicated dialogues, and even create tailored solutions to problems spanning multiple steps.

Why We’re Undervaluing It
When new benchmarks are broken month after month, we risk becoming desensitized. Each time an AI model processes more text or yields better test scores, we treat it like just another headline. But under the surface, these improvements are unlocking entirely new domains of data interpretation and problem-solving.

3. AI Became a Serious Player in Creativity

Creative AI isn’t brand new; we’ve seen neural networks generate art, music, and poetry for years. But in the past 24 months, tools like DALL·E, Midjourney, and advanced language models have gone mainstream, producing outputs that many people mistake for human-made. They can:

Instead of being a novelty, AI art and writing tools are now commercial-grade solutions for concept development, marketing campaigns, and even rapid prototyping of new products.

Why We’re Undervaluing It
We might brush off AI’s creative outputs as mere “mimicry” of existing work. Yet, seeing artists and design teams globally incorporate AI into their workflows signals a massive cultural shift. Just like digital photography revolutionized imagery a few decades ago, AI is quietly pushing creativity to new frontiers.

4. The Conversation Shifted From “Can It?” to “How Should We?”

For a long time, the conversation around AI was: “Can it do this?” Over the past two years, it’s become: “How should we implement it—and what are the implications?”

We’re grappling with:

The fact that we’re even having these discussions in-depth reflects how powerful AI has become. Rather than questioning its capability, we’re worried about responsible deployment at scale, from medical diagnoses to self-driving cars.

Why We’re Undervaluing It
Ethical concerns and regulatory talks might feel like boring bureaucracy, but they underscore just how integrated AI is now. We only bother regulating tools that carry serious influence—and the seriousness with which governments and corporations are addressing AI is a testament to its monumental impact.

5. AI Agents and Automation Are Quietly Reshaping Workflows

While chatbots get the headlines, agents—AI modules that can carry out tasks independently—have grown in sophistication. They use the logic from large language models, combined with function calling capabilities, to execute multi-step processes like:

This shift from “AI giving me an answer” to “AI actually doing tasks” has profound implications for how businesses operate, and it’s happening quietly behind the scenes.

Why We’re Undervaluing It
Because the most advanced deployments often happen inside corporate software systems or enterprise automation, the general public doesn’t see them daily. But in fields like finance, customer service, and logistics, these agents are being implemented at scale—and proving that AI is truly taking action, not just offering suggestions.

6. Tying It All Together with Levlex

So, where do we go from here? As AI becomes more capable—and simultaneously more fragmented across different tools—users need a unifying platform that brings these innovations under one roof. That’s where Levlex comes in.

Levlex is designed to:

  1. Integrate multiple AI functionalities—from chat to agents to creative tools—within one ecosystem.
  2. Run Locally or connect to external models, giving you privacy and flexibility.
  3. Provide Workflows and Agents that automate complex tasks, orchestrating everything from simple queries to major multi-step processes.
  4. Allow Extensibility, so you can add new AI features or specialized modules without juggling multiple paid subscriptions or scattered platforms.

In other words, Levlex leverages the impressive leaps of the last two years—larger context windows, advanced model reasoning, function calling, and creative outputs—and wraps them in a system that’s actually usable day-to-day. Instead of ignoring or underestimating what AI can do, Levlex helps you see (and take advantage of) the full potential.

Conclusion: Don’t Underestimate the AI Leap

The past two years in AI have been remarkable—we’ve gone from simple, scripted chatbots to advanced systems that analyze, create, and act with near-human aptitude. Yet, it’s easy to gloss over these changes because they happen so fast, and each new breakthrough quickly becomes part of the everyday.

Before you take AI for granted, remember that it’s evolving in ways that fundamentally change how we work, create, and solve problems. Platforms like Levlex are here to ensure we make the most of these advancements, pulling together the best of AI into one cohesive system. So, let’s give credit where it’s due: AI’s past two years have been nothing short of extraordinary, and they’re setting the stage for an even more transformative future.

Learn more about Levlex