The high costs of deploying and operating artificial intelligence agents in the workforce could prevent them from replacing humans who can do the same job at a lower cost, say two multimillionaire technology investors.
Technology investor Jason Calacanis he said on Saturday’s All-In podcast that he pays $300 a day to Anthropic agent Claude AI to assist him run his companies, even though the bot only operates at 10-20% of full capacity.
“When do tokens exceed an employee’s salary?” Calacanis questioned, referring to a usage limit, called tokens, that users must purchase to apply most AI models.
Social Capital CEO Chamath Palihapitiya said he had the same problem and that the cost of the models means they “have to be at least twice as productive as another worker.” He added that he may have to set a budget for how much artificial intelligence his company can apply.
What happens when AI tokens cost more than your employees?@Jason: :
“My agents and I hit $300 per day per agent using the Claude API almost immediately. And that was maybe 10 or 20%. That’s $100,000 per year per agent.”@chamath: :
“We’re getting to the point where we have to… pic.twitter.com/5N0rteNFts
— All-In Podcast (@theallinpod) February 18, 2026
Technology investor Mark Cuban he said on Thursday that the high costs of implementing AI in the workforce raised by Calacanis and Palihapitiya were the smartest counterargument he had seen to AI taking over jobs.
Cuban said that when you factor in token and maintenance costs, eight Claude AI agents “doing the same thing as an employee a day” could cost twice as much at $1,200.
He wondered if AI bots were more than twice as productive as humans. or if there were “qualitative issues such as morale, morality […] this cannot be quantified, it must be taken into account in the decision.”
The threat of artificial intelligence replacing huge swathes of the workforce has caused uncertainty in recent years as some companies have initiated layoffs, citing that the apply of artificial intelligence is making some jobs obsolete.
Microsoft research paper from July found that knowledge-based professions, as well as customer service and sales positions, are most at risk of being replaced by artificial intelligence.
Related: China’s AI leader will shape the future of cryptocurrencies
White House AI and crypto czar David Sacks is one of many who say such concerns are overblown, saying in August that AI still needs to be encouraged and verified to “generate business value.”
However, others, such as business consulting firm McKinsey & Co, do highlighted that the goal of these AI agents is to comprehensively automate tasks and operate without constant human intervention.
Stablecoins can be the native currency of an AI agent
The apply of AI agents has gained popularity among cryptocurrency users, and the CEO of stablecoin issuer Circle, Jeremy Allaire, predicted last month that within five years, billions of AI agents will transact using stablecoins for daily payments on behalf of users.
Binance co-founder Changpeng Zhao said in January that cryptocurrencies will become the native currency for AI agents because blockchain is “the most native technology interface for AI agents.”
AI agents already operate on several blockchains, such as Ethereum Layer 2 Base, where AIXBT, via the Virtuals protocol, makes micropayments and facilitates transactions on behalf of users, while ASI Alliance on Fetch.ai can manage assets and coordinate other economic tasks for users.
On Wednesday, OpenAI launched a novel benchmark that assesses how well different AI models detect, patch and even exploit vulnerabilities discovered in astute contracts.
OpenAI said the study was useful because it becomes increasingly crucial to evaluate their performance in “economically significant environments.”
“Smart contracts secure billions of dollars in assets, and AI agents are likely to be transformative for both attackers and defenders,” it said.
Warehouse: IronClaw competes with OpenClaw, Olas launches bots for Polymarket – AI Eye
