The call came in from Dubai while Peter Diamandis was making breakfast for his kids. Diamandis was hosting an audience question round of Moonshots with Peter Diamandis. The episode was published on September 22, 2026. The caller, identified on air only as GS, said he had spent 20 years in the textile and apparel export industry. He wanted advice on how to take his AI company forward, so that the factories he knows could compete instead of "going out of market."
A plan built on protecting jobs
GS said the idea came from the show itself. He heard the hosts talk about the "organizational singularity," in which AI pushes transaction costs toward zero. Transaction costs are the costs of coordinating business: drafting and checking documents, arranging orders, negotiating and keeping records. For him, that idea "was the ring bell." He pointed to labor issues in the part of the world where he works. If digital tools and AI make transaction costs collapse, he argued, his industry has to shape its own future now.
GS used the show's term MTP, short for "massive transformative purpose," for his guiding mission. He said his MTP is to reach "multi-thousand factories" as well as buying houses and trading houses. Buying houses and trading houses are the middlemen that connect factories with overseas buyers and handle their orders. He wants all of them to use AI and compete globally. Protecting workers is central to that goal. Millions of people work in the industry in his region, he said. Saving those jobs means making the white-collar processes around the factory floor "agentic," so that efficiency keeps improving.
His company, whose name the episode transcript renders as Partham.ai, plans to build that agentic layer. AI agents are software that carries out tasks rather than just answering questions. The target customers are small enterprises that would work this way without expensive software and ERP systems. ERP, or enterprise resource planning, refers to the large software suites that big companies use to track orders, inventory, production and finances.
The size of the problem, as GS tells it
GS gave one number to show why this matters. In his part of the world, he said, transaction costs in the industry amount to 10 to 20% of sales. By his account, that is five to ten times what the industry earns in profit. On his figures, the cost of coordinating the work far exceeds what the factories make. That is the scale of inefficiency he wants to go after. The figure is his own estimate of his industry. He did not say how it was measured.
Dave Blundin: skip the robots, start with the office
GS put his question to two of the show's co-hosts, Dave Blundin and Salim Ismail. Blundin said he was "so, so glad" GS had asked, because he sees the same pattern often. Manufacturing is considered labor-intensive, he said. But when you look at a manufacturer's actual operating costs, it is all about "planning transactions, documents, all of which is beautiful AI territory."
When he looks at the numbers, Blundin said, the reaction is that AI could drop "10, 20, 30 % more to the bottom line with stuff that AI can do right now." This is a separate figure from GS's. GS was describing how much of a company's sales goes to transaction costs. Blundin was giving a rough guess at the extra profit AI might produce. Neither speaker cited a deployment where those gains were measured.
Blundin also rejected a common assumption. "Everyone's thinking, oh, but first I would need to, you know, have robots all over the place," he said. "Like, no, no, no, no." Instead, he said, start with the paperwork, planning and scheduling, and with "the inefficiency of where people are and what they're doing": how staff are assigned across the work. He called all of that "perfectly attackable with AI."
His advice on building the business was to productize the workflow. That means turning it into a standard product sold across a region to one class of companies, rather than custom work for each client. Done that way, he said, it "should work incredibly well," and he called it "a very, very cool idea."
Where to launch first
Blundin then made a broader point about where AI will be adopted. Many people expect all wealth and power to go to two or three places in the world, he said. But he believes regulation will make many things "very, very slow" in those places, including the United States and California. He said this happens often in biotech and predicted the same for self-driving cars and delivery drones. Those technologies, he said, will spread faster in other parts of the world.
His suggestion was to look for areas where a technology could run in China or the US but is held back by government action. Then deploy it wherever it "can take off and flourish," and bring it back to the US later. He said that has "always been a really good business plan." Blundin offered it as a general strategy and did not say how it would apply to GS's industry specifically.
Blundin then passed the question to Ismail, asking whether he was in the middle of some big Bangalore deal. Ismail said he was just trying to make his way through the city and declined to add anything.
A different use for cheaper coordination
The organizational-singularity idea that inspired GS is usually about how cheaper coordination could change why companies exist and where their boundaries lie. GS is aiming at something more concrete. He wants cheaper coordination to help thousands of existing factories and their middlemen survive global competition, with their workforces intact. For now, what he described on the show is a plan, and the savings he and Blundin talked about are their estimates rather than results from factories using the system.