Humain pushes for an AI-first computing experience — but there are skeptics

Five months ago, Humain made a bold promise: it would change the way people interact with computers.

The Saudi Arabian company, launched only last May, argued it could replicate the human experience on computers, especially for work chores. AI Agents, Humain said, could replace apps and take instructions automatically to handle various work tasks.

After launch, the company quickly got to work on building a full stack of the computing model. The base is a home-grown operating system with a barebones interface built atop AI agents. Using that OS, customers can type in a command or verbally explain what they want done; the agents then follow up and complete the tasks while people deal with other duties. The OS essentially does away with the icons or click-throughs common in other modern-day OSes.

“I am 100% convinced — whether Humain does it, Google does it, Apple does it —  this is the future UX,” Humain CEO Tareq Amin said at the FII9 conference last month in Riyadh. (He was part of a fireside chat alongside Alphabet President Ruth Porat and Oracle CEO Mike Sicilia.)

“We will get into a world where AI agents are integrated into one platform,” Amin said. “They are directly integrated with your enterprise system processes.”

Potential rivals like Microsoft and Google are already hyping their own AI-infused operating systems and productivity suites. Microsoft has made its Copilot tool an integral part of Windows 11, while Gemini is a key part of Google’s cloud-native tools. On the browser front, OpenAI and Perplexity have both shipped AI-centric  browsers.

As for Humain, its ambitious OS can go into a homegrown laptop built in partnership with Qualcomm, which will supply Snapdragon chips; the last piece of the stack are Humain data centers spread across Saudi Arabia. Those facilities, with gigawatts of capacity, would provide the computing capacity for AI agents to do their tasks.

A Humain Horizon Pro PC.Mark Hachman / Foundry

Humain bought GPUs from Nvidia and AMD for the data centers, with Amin appearing on stage at events with Nvidia CEO Jensen Huang and AMD CEO Lisa Su to tout the partnerships.

With the various parts in place, the company last week unveiled Humain One, which unifies  the whole computing stack.

“With Humain One, we are redefining enterprise computing by making AI an active partner that understands your goals, anticipates needs, and executes tasks autonomously.,” Amin said in a statement. “Moving from cluttered icons to streamlined intelligence, this is an adaptive intelligence layer built to elevate productivity and creativity across every role in the enterprise.”

The company was unable to immediately respond to questions from Computerworld about its plans.

A (computing) change is coming…

Desktops have relied on icons going back to the early days of Windows in 1985, but it’s time for that UI to end, Amin said.

“You’ll end up with thousands of applications that organizations are built around it. And then we wonder why value realization is not delivered,” he said.

With AI agents making it possible to realize value in enterprises, it’s time to take advantage of that technology for tasks such as HR and finance, Amin said. “With advanced models that Google has done, OpenAI has done, I pondered why the interfaces on mobile computers never changed,” Amin said.

Humain’s plan is ambitious and could change the way companies think about AI-driven interaction, analysts said. With the notion that traditional apps could disappear and be replaced by agents working in a new type of environment, they agreed Humain has an opportunity.

“The question is: are they the ones that are going to be able to pull that off? That’s not quite as clear,” said Bob O’Donnell, principal analyst at Technalysis Research.

Humain OS shown doing research.Mark Hachman / Foundry

Despite some skepticism, analysts noted that Humain is funded by Saudi Arabia’s Public Investment Fund, which has close to $1 trillion in funding. Beyond that, Saudi giant Aramco, which wants to convert energy into tokens in the data centers, recently took a stake in the company. (The tokens could create AI-driven revenue.)

“We need to take Humain seriously, given the backing,” said Jack Gold of J. Gold Associates. “But how far they can push the envelope remains to be seen.”

Enterprise users have seen glimpses of AI in browsers and productivity software already, but an AI-powered OS is a completely different challenge. “There is no doubt we will see increasingly agentic AI interfaces that — if not totally doing away with the OS — at least [have] a major disruption of the UI,” Gold said.

Humain’s early deployment involves Arabic large language models (LLMs) for the Middle East region. But the scale of their deployment points to global ambitions. 

Saudi Arabia, like the UAE, wants to become a global AI powerhouse with heavy government backing, analysts said.

“All the other major AI organizations are pursuing a similar path — and the Chinese are not exactly shy on resources to make it happen,” Gold said.

Pushing rivals ‘out of their comfort zone’

Humain’s idea of replacing apps with intent will be interesting to watch, said Stephanie Walter, practice leader of the AI stack at consulting firm Hyperframe Research. “I really do think Humain is going to push the incumbents out of their comfort zone,” she said.

Some early Humain partners include EY for HR, tax, finance and other services, and Replit for software development in the region.

Now, Humain needs to convince software companies that it is legitimate and they should spend money making apps compatible with the Humain OS.  “As other OS makers catch up or adopt their own operating systems to AI, it’s going to be tricky,” Walter said.

Enterprises will need to worry about security and guardrails around the AI agents and where any corporate data might be stored. That could become a major issue for companies operating outside the US.

Additionally, there will likely be concerns about how much to trust data outputs, or whether hackers might have interfered with user prompts. In many cases where autonomous agents are already executing processes, enterprises still have humans in the loop.

“They have to worry about the quality and accuracy of their data…and what LLMs are putting out…,” Walter said. “That’s why some of these AI projects are being approached rather cautiously.”

Humain’s success will come down to adoption and execution, she said. But the company has a long runway for success.

“They’ve covered the funding and don’t have to worry about revenue for a long time,” Walter said.Humain pushes for an AI-first computing experience — but there are skeptics – ComputerworldRead More