The high costs of deploying and running artificial intelligence agents in the workforce may prevent them from replacing humans who can do the same work at lower cost, say two multimillionaire tech investors.
Tech investor Jason Calacanis said on the All-In podcast on Saturday that he has been paying $300 per day for an Anthropic Claude AI agent to help run his businesses, despite the bot only operating at 10% to 20% of full capacity.
“When do tokens outpace the salary of the employee?” Calacanis questioned, referring to the usage allowance, called tokens, that users must purchase to use most AI models.
Social Capital CEO Chamath Palihapitiya said he had the same problem and that the cost of the models means they “need to be at least two times as productive as another employee.” He added he may need to set a budget on how much AI his business can use.
What Happens When AI Tokens Cost More Than Your Employees?@Jason:
— The All-In Podcast (@theallinpod) February 18, 2026
“We, with our agents, hit $300/day per agent using the Claude API, like instantly. And that was doing, maybe, 10 or 20%. That's $100k/year per agent.”@chamath:
“We're getting to a place where we have to… pic.twitter.com/5N0rteNFts
Tech investor Mark Cuban said on Thursday that the high cost of AI adoption in the workforce raised by Calacanis and Palihapitiya was the smartest counter-argument he’d seen to AI taking over jobs.
Cuban said that with the cost of tokens and maintenance, it could cost twice as much for eight Claude AI agents “to do what an employee does per day” for $1,200.
He questioned whether the AI bots were more than twice as productive as a human. or if there were “qualitative issues like morale, morality […] that can’t be quantified, that need to go into the decision.”
The threat of AI replacing large swathes of the workforce has caused uncertainty in recent years, as some companies have initiated layoffs, pointing to their use of AI making some jobs obsolete.
A research paper from Microsoft in July found that knowledge-based occupations, along with customer service and sales roles, were most at risk of being replaced by AI.
Related: China’s AI lead will shape crypto’s future
White House AI and crypto czar David Sacks is one of many who claim such fears are overhyped, saying in August that AI still needs to be prompted and verified to “drive business value.”
However, others, such as business consulting firm McKinsey & Co, have highlighted that the point of these AI agents is to automate tasks end-to-end, operating without constant human input.
Stablecoins could be agentic AI’s native currency
The use of AI agents has grown in popularity among crypto users, and stablecoin issuer Circle CEO Jeremy Allaire predicted last month that billions of AI agents will be transacting with stablecoins for everyday payments on behalf of users within five years.
Binance co-founder Changpeng Zhao said in January that crypto would end up being the native currency for AI agents due to blockchain being the “most native technology interface for AI agents.”
AI agents are already operating on several blockchains, such as Ethereum Layer 2 Base, where AIXBT, via the Virtuals Protocol, makes micropayments and facilitates trades 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 new benchmark evaluating how well different AI models detect, patch and even exploit security vulnerabilities found in smart contracts.
OpenAI said the research was useful as it is becoming more important to evaluate their performance in “economically meaningful environments.”
“Smart contracts secure billions of dollars in assets, and AI agents are likely to be transformative for both attackers and defenders,” it said.
Magazine: IronClaw rivals OpenClaw, Olas launches bots for Polymarket — AI Eye
