Smart AI Is Here: What’s Next for Tech?
by Samson Odo, Co-Founder / CTO
A Chinese startup, Monica just unveiled something truly impressive: an AI called Manus that tackles tough tasks solo. It can conduct PhD-level research, track stock trends, or even build websites and games without someone guiding it every step of the way. Announced in early March 2025, it’s already a big deal in the tech world, sparking chatter — are we getting close to AI that thinks like a human?
Manus is still in testing, but within a week of its announcement, some clever users have already peeked at its setup and started making free versions anyone can use.
That’s part of a bigger shift: high-end AI isn’t just for the pros anymore — it’s going mainstream fast, with open-source versions popping up almost overnight.
Tech Hits a Wall
Think about smartphones. Early models were game-changers—touchscreens, apps, the works. Remember the jump from flip phones to the first iPhone?
Now? It’s mostly just tweaks like a better camera or longer battery life. Tech often starts with a bang and then settles down. AI is undoubtedly heading in that direction too. The giant leaps in speed and smarts—like when AI first beat humans at chess or Go—are slowing, and what’s next might feel more like small upgrades.
The real edge for AI won’t be about raw power. It will come from how it’s built for specific tasks, how simple it is to use, and what data it taps into. The best AIs will slide right into your work or life, solving problems without making things complicated—like a virtual assistant that just gets you.
Top tip
Experts dug into Manus and found it’s not about fancy new tech. It’s a clever mix of existing models, powered up by custom prompts, smart tool use, slick workflows, and sharp evaluations — proof today’s AI is already plenty smart if you use it right.
Sound Familiar?
All this AI hype might remind you of the late ‘90s dot-com boom. Back then, investors poured money into random internet startups, convinced they would change the world—until the bubble burst in 2000.
Today’s AI scene, with its wild excitement and big investments, seems to be on a similar path. And it’s not just talk-those AI data centers are using so much electricity that Texas’s power grid is straining to keep up, with demand spiking in 2024 alone. It’s a heads-up that even awesome tech has limits, from cash to kilowatts.
AI Becomes Invisible
Soon, AI will be as ordinary as the phone in your hand. You don’t think about how your phone works—you just use it.
AI is heading there too, quietly slipping into your job, shopping, and daily routine. Scheduling meetings, suggesting purchases, or optimizing your commute without a second thought. It’s a huge change, but once it’s everywhere, it won’t seem like a big deal anymore.
Where This Is Going
The days of AI wowing us with massive breakthroughs are winding down. Now, it’s about smarter, practical steps forward—like making tools that fit real needs, not just chasing headlines.
At Nodeshift, we’re not chasing the hype of building our own foundation models. Why start from scratch when existing ones, like those powering Manus, still have so much potential?
We believe the real game-changer is in the tooling and integration—unlocking what’s already there with a sharp, systems-thinking approach.
That means looking at the big picture: how AI connects with workflows, data, and people to solve problems in ways off-the-shelf tools can’t touch.
The difference won’t be between companies with generic AI tools and those without. It will be between those who just plug in basics and those who weave AI into business and society in fresh, powerful ways.
Custom solutions that boost efficiency and spark new ideas — not just chatbots reciting scripts.