Welcome back, AI Enthusiasts!
In today’s AI summary rundown:
On December 30, 2025, Meta announced it would acquire Manus, a Singapore-based AI startup known for its autonomous, general-purpose AI agent capabilities, integrating them into Meta AI and other products.
The acquisition’s valuation is estimated between US$2–3 billion, although Meta has not formally disclosed financial terms. Manus had quickly achieved an annual recurring revenue (ARR) exceeding $100 million within eight months of its March 2025 launch and reportedly reached a total revenue run rate of over $125 million.
Manus garnered attention earlier in 2025 for demonstrating a general AI agent capable of executing multi-step tasks such as market research, coding, and data analysis autonomously, performance it claimed exceeded that of OpenAI’s DeepResearch. The platform has processed over 147 trillion tokens and spun up 80 million virtual computers to date.
Meta intends to let Manus continue operating as a standalone service from its Singapore base while integrating its technology into Meta’s apps and tools. Amid U.S.-China geopolitical sensitivities, Meta confirmed Manus will fully sever its Chinese ownership ties and cease operations in China.
This acquisition marks Meta’s third-largest deal, after WhatsApp and Scale AI, in its aggressive AI-driven expansion strategy. The acquisition reinforces Meta's ambition to embed task-oriented ‘agentic’ AI systems into products across its ecosystem and stay competitive with AI leaders like OpenAI and Google.
Read time is 4 min..
AI Insights
Wedge Issue Alert: AI Emerges as a Divisive Political Flashpoint
Polling reveals widespread public negativity toward AI growth
Surveys from Gallup and Pew, highlighted by Political; Politico, show that a solid majority of U.S. voters regard the rapid rise of AI as a net negative development.
Growing public demand for AI regulation
Americans are increasingly advocating for government oversight of AI. According to data from Pew, majorities in both political camps fear government regulation will be too lax, but still express little confidence in their ability to implement it effectively.
Public anxiety sharply outweighs optimism
Pew surveys report that most Americans are more concerned than excited about AI; around 50–57% express stronger worry, with optimism lagging significantly.
A deep divide between public and AI experts
While the general public is apprehensive, close to 75% of AI experts remain optimistic, underscoring a widening trust gap.
Across-party concern signals AI’s potential as a wedge issue
The alignment of both MAGA populists and Democrats in expressing caution toward AI development signals its emergence as a crossover wedge issue, capable of shaping voter coalitions.
AI Training: AI Tutorial of the Day
Turn the Rise of AI Agents Into a Clear, Practical AI Brief
One thing I’ve been experimenting with on thareja.ai is understanding what people actually mean when they talk about AI agents, beyond the hype and buzzwords.
This is where testing different models really highlights their strengths.
Start a New Chat
Open thareja.ai and start a new conversation.
I like to isolate one emerging concept per chat so the thinking stays focused.
Switch Your AI Model
Click the ( + ) icon next to Automatic and choose your model.
For this tutorial, I selected GPT-4o.
Why GPT-4o?
It’s particularly strong at:
Breaking down complex systems
Explaining workflows step by step
Translating technical ideas into business relevance
Try This Prompt & Observe the Difference
Prompt used:
Explain what AI agents are in simple terms.
Describe how they differ from traditional chatbots.
Give one practical example of how a business might use them today.
Keep it under 150 words.
Model Used
GPT-4o
AI Response (Excerpt)
AI agents differ from chatbots in that they don’t just respond; they act.
They can plan tasks, make decisions across steps, and interact with tools or systems without constant human input.
For example, a sales operations agent might monitor inbound leads, enrich data, assign priority, and schedule follow-ups automatically.
Why This Experiment Works
When I ran the same prompt on Claude, the explanation became more conceptual and abstract.
GPT-4o delivered a clearer, execution-focused explanation.
This is exactly why I enjoy using thareja.ai, switching models instantly makes emerging concepts easier to evaluate without getting lost in hype.
Happy Prompting!
Exclusive Member Deal - 20% OFF
Superpower AI Bundle Access 50+ Major LLMs
One Subscription. $20/mo $16/mo with code SUPERPOWER20
AI-Generated Image of the Day
Prompt: Create a hyper-realistic cinematic image of Amelia Earhart in a futuristic airport hangar, examining a sleek, advanced aircraft. Dramatic lighting highlighting her pioneering spirit, detailed textures on the aircraft, and a bustling futuristic airport in the background. Ultra-HD resolution, conveying adventure and innovation.
Tip: the more specific the better
What I like most about thareja.ai is the freedom to switch models quickly and see creativity take new shapes. Each model offers something unique, but Nano Banana is my go-to for bold, energetic imagery. Have fun exploring what each one can do.
Meme of the Day
Question of the Day
Why is human-in-the-loop design important for critical AI applications?
That’s it for today’s news in the world of AI!
If you have anything interesting to share, please reach out to us by sending us a DM on Twitter: @dthareja or email me at [email protected]. How was today's newsletter?
Feedback helps us improve!
Thanks for reading. Until next time!
p.s. if you want to sign up for this newsletter or share it with a friend or colleague, you can find us here.