Looks like they only included actual chats and not agentic/copilot usage. IMO that makes the study quite incomplete.

The chats alone are backbreakingly costly relative to the market mix of ChatGPT.

Rest of the market be damned -- combined with the poor customer mix (low to middle income countries) this explains why there has been such a push by the big labs to attempt to quantize models and save costs. You effectively have highly paid engineers/scientists running computationally expensive models on some of the most expensive hardware on the market to serve instructions on how to do things to people in low income countries.

This doesn't sound good, even for ad-supported business models.

	▲	haijo2 17 hours ago \| parent \| next [-]
		Yep and lets not forget, those people are incredibly price sensitive. Is there enough product differentiation between OAI and Gemini? Not that I can see. And even if it was a low price, thats not the point - people hate paying a penny for something they expect to be free. By the time OAI has developed anything that enables them to acquire and exercise market power (profitably), they will have ran out of funding (at least on favourable terms). Which could cause key talent to leave to competitors and so on. Essentially a downward spiral to death.
	▲	ares623 20 hours ago \| parent \| prev [-]
		I also wonder how much of those "writing" assistance is for propaganda, troll farms, or scams. Such value. $500B well spent.