This was a pretty awesome read. The charts of people using AI for search are mind-blowing but not at all surprising. Using AI to create content is one thing. Using it to streamline how efficiently you operate is the next level up.
And as somebody who helps people build brands and operating systems, this acquisition makes Manus a very attractive option.
This is the smartest $2B Meta's spent in years. It's like watching everyone race to build faster engines while Meta just bought the only car that's actually driving people to work.
So true "the Platforms decide on "your" AI" 🤓 when AI works in the workflow it's NOT called AI anymore it's simply called software, the system, the workflows... only new AI is called AI 🌊
I think 2026 will see a greater adoption, as you guys shown from the graph in the article. More push for autonomous systems, agents, subagents, and that kind of direction. We shall see!
I actually think 2026 will be about consolidation of cash-cow services, not technological breakthroughs. Companies will focus on amassing money, not advancing the technology base.
AI winter is coming. The LLM investment curve has flattened while costs keep exploding. The next 18 months will separate companies with real revenue models (like Manus) from those burning cash on capability demos.
Meta's move is defensive positioning before that winter hits. They're locking in distribution and revenue NOW, while others are still chasing the next model improvement
Hm, that’s an interesting take, I kind of see it though. Meta really slowed down their development of their models, but now they definitely pivoted in their strategy.
"The question isn’t whether AI will change your work. The question is which company’s AI will be making those changes, and whether you’ll understand what’s happening in time to guide it strategically."
Well said. Time to get your arms around the strategy.
Exactly. Business is strategy, technology is just one lever to pull. Meta isn't committed to building the best AI. They're committed to growth and stakeholder promises. Manus delivers both: immediate revenueand a path to defend their market position against the next paradigm shift. Most companies get this backwards. They chase technological superiority while competitors are executing strategic positioning. By the time the 'better technology' ships, the market is already locked in.
Good stuff. I had originally stayed away from manus after a brief testing period because I was afraid of where my data was going (to your point on yesterdays live). But in my testing I found it incredibly powerful
Exactly, and that's the real story. Their architecture is the moat, not just the model. Ranking at 86% in reasoning puts them right alongside Gemini and ChatGPT, but what matters is how they got there.
Manus built agent-native from the ground up: persistent execution loops, task graphs, long-term memory, tool orchestration as first-class. That's not something you retrofit onto an LLM wrapper.
I had not jumped to Manus yet, that’s why I wanted @Farida to write this guest post. I have been very disappointed at Metas AI work so I hope Manus doesn’t start going downhill now
Thank you for this brilliant article Joel and Farida. This is by far the most insightful take on the Manus acquisition that I have read. Going to be really interesting to see how they turn it lose on all that data, and the extent to which the decide to apply ethical considerations at all...
Unfortunately, Meta's playbook is about monopoly, aggressive expansion with minimal privacy considerations. That's the pattern from Instagram, WhatsApp, and now Manus.
The interesting wrinkle: the deal includes keeping Manus's management and team intact. Whether that provides any ethical guardrails or just gives Meta plausible deniability while they optimize for scale... we'll see soon enough.
My guess: the Manus team stays autonomous until integration priorities conflict with their values. Then we'll learn whether they have real influence or were just part of the acquisition package
Sam, I echo every single one of those sentiments. What happens when a solid tool is integrated into an organization with lackluster guardrails and a poor track record of safety?
One thing that really landed and resonated with me is this: over the next couple of years, the execution gap will be the real divide between everyday and power AI users. I’m really curious to see how everyone here adapts.
Great deep dive into Manus, team! Quite insightful as to why Meta would do this acquisition rather than built in-house. Curious to try out Manus AI tool again once it is 'meta-morphosed'!
What'll be interesting is whether Meta keeps Manus as a standalone product or absorbs it completely into Meta AI. My bet: they'll do both, keep the Manus brand for enterprise/power users while embedding the core architecture into Instagram and WhatsApp for mass distribution.
Try it before the integration if you can, you'll get a sense of what the 'pure' Manus architecture feels like versus what it becomes once Meta optimizes for scale and monetization. Could be very different products in 6-12 months.
Exactly, and that's precisely why Meta moved so fast. Most people still don't realize how powerful Manus is, which gave Meta a narrow window to acquire before the market fully understood its value
Great breakdown!
Will be interesting to see where this lands in Meta's portfolio on the Whatsapp to metaverse scale
we will see integration with Instagram first I guess, then WhatsApp Business
As much as they’ve been struggling with a good product to compete with, I would hope so
I’d me curious to see that too
Meta will become aggressive in 2026 just to catch up in the AI race, mostly about presence not technology
This was a pretty awesome read. The charts of people using AI for search are mind-blowing but not at all surprising. Using AI to create content is one thing. Using it to streamline how efficiently you operate is the next level up.
And as somebody who helps people build brands and operating systems, this acquisition makes Manus a very attractive option.
Thanks for the great write-up
Farida did a great job! I don’t know that’s what you did, Liam. That’s awesome! And a field where I think demand is just going to grow.
Clearly I need to get better at the branding part, huh?
Haha it’s good to know! I’ll keep it in mind for a future collab if you are open
Thank you Liam, happy it resonates
This is the smartest $2B Meta's spent in years. It's like watching everyone race to build faster engines while Meta just bought the only car that's actually driving people to work.
great analogy!
Meta always had a hawkish business attitude
Great post!
I wonder how you think Meta’s reputation might affect this acquisition and its long-term potential.
Personally, I find it hard to trust anything that comes directly from Meta or Zuckerberg’s vision.
Trust is essential for leadership — and that’s why I’d love to hear your perspective on this.
So true "the Platforms decide on "your" AI" 🤓 when AI works in the workflow it's NOT called AI anymore it's simply called software, the system, the workflows... only new AI is called AI 🌊
Only new ai is called ai now, so true!!
The capability gap isn’t about models anymore, it’s about how intentionally leaders use them.
Yup! That's why I'm curious to see what a company that has made very dubious AI moves does with a well-run acquisition haha
Will be a the year of strategies and execution
I really love this breakdown, especially the insights into the architecture and the importance of the agentic layer
Really interested to see how this impacts their market position 6 months down the line
Great post guys!
2026 will be year of too much gas actions
Amazing article you guys.
I think 2026 will see a greater adoption, as you guys shown from the graph in the article. More push for autonomous systems, agents, subagents, and that kind of direction. We shall see!
Appreciate it Ilia
I actually think 2026 will be about consolidation of cash-cow services, not technological breakthroughs. Companies will focus on amassing money, not advancing the technology base.
AI winter is coming. The LLM investment curve has flattened while costs keep exploding. The next 18 months will separate companies with real revenue models (like Manus) from those burning cash on capability demos.
Meta's move is defensive positioning before that winter hits. They're locking in distribution and revenue NOW, while others are still chasing the next model improvement
Hm, that’s an interesting take, I kind of see it though. Meta really slowed down their development of their models, but now they definitely pivoted in their strategy.
Such a key point right here:
"The question isn’t whether AI will change your work. The question is which company’s AI will be making those changes, and whether you’ll understand what’s happening in time to guide it strategically."
Well said. Time to get your arms around the strategy.
Exactly. Business is strategy, technology is just one lever to pull. Meta isn't committed to building the best AI. They're committed to growth and stakeholder promises. Manus delivers both: immediate revenueand a path to defend their market position against the next paradigm shift. Most companies get this backwards. They chase technological superiority while competitors are executing strategic positioning. By the time the 'better technology' ships, the market is already locked in.
Thank you, Patrick! Such an important question but one so few leaders are asking
Good stuff. I had originally stayed away from manus after a brief testing period because I was afraid of where my data was going (to your point on yesterdays live). But in my testing I found it incredibly powerful
Exactly, and that's the real story. Their architecture is the moat, not just the model. Ranking at 86% in reasoning puts them right alongside Gemini and ChatGPT, but what matters is how they got there.
Manus built agent-native from the ground up: persistent execution loops, task graphs, long-term memory, tool orchestration as first-class. That's not something you retrofit onto an LLM wrapper.
I had not jumped to Manus yet, that’s why I wanted @Farida to write this guest post. I have been very disappointed at Metas AI work so I hope Manus doesn’t start going downhill now
lets see! Maybe this will be the thing that turns the ship around~
Thank you for this brilliant article Joel and Farida. This is by far the most insightful take on the Manus acquisition that I have read. Going to be really interesting to see how they turn it lose on all that data, and the extent to which the decide to apply ethical considerations at all...
Unfortunately, Meta's playbook is about monopoly, aggressive expansion with minimal privacy considerations. That's the pattern from Instagram, WhatsApp, and now Manus.
The interesting wrinkle: the deal includes keeping Manus's management and team intact. Whether that provides any ethical guardrails or just gives Meta plausible deniability while they optimize for scale... we'll see soon enough.
My guess: the Manus team stays autonomous until integration priorities conflict with their values. Then we'll learn whether they have real influence or were just part of the acquisition package
Sam, I echo every single one of those sentiments. What happens when a solid tool is integrated into an organization with lackluster guardrails and a poor track record of safety?
One thing that really landed and resonated with me is this: over the next couple of years, the execution gap will be the real divide between everyday and power AI users. I’m really curious to see how everyone here adapts.
Adaptation isn't optional, it's survival. You either get on board or get left behind
Great deep dive into Manus, team! Quite insightful as to why Meta would do this acquisition rather than built in-house. Curious to try out Manus AI tool again once it is 'meta-morphosed'!
Haha, love the 'meta-morphosed' pun
What'll be interesting is whether Meta keeps Manus as a standalone product or absorbs it completely into Meta AI. My bet: they'll do both, keep the Manus brand for enterprise/power users while embedding the core architecture into Instagram and WhatsApp for mass distribution.
Try it before the integration if you can, you'll get a sense of what the 'pure' Manus architecture feels like versus what it becomes once Meta optimizes for scale and monetization. Could be very different products in 6-12 months.
That's good advice, Farida! Only time will tell if and how Manus architecture changes with this deal. Curious to try it out in the coming days!
The model is really good , I have been using them for sometimes
I didn’t know they bought Manus! It’s one of the most powerful and most underappreciated tools in AI right now.
Exactly, and that's precisely why Meta moved so fast. Most people still don't realize how powerful Manus is, which gave Meta a narrow window to acquire before the market fully understood its value