The AI industry loves token inflation. Your company shouldn’t
The AI industry has a quiet addiction problem: It is addicted to tokens. Every new generation of agentic AI seems to assume that the answer to complexity is to throw more context at the model, keep longer histories, spawn more calls, loop over more tools, and let the token meter run wild. The rise of agentic systems, and now projects like OpenClaw , makes that temptation even stronger. Once you give models more autonomy, they do not just consume tokens to answer questions. They consume them to plan, reflect, retry, summarize, call tools, inspect outputs, and keep themselves on track. OpenClaw itself describes the product as an “agent-native” gateway with sessions, memory, tool use, and multi-agent routing across messaging platforms—which tells you exactly where this is going: more autonomy
The AI industry has a quiet addiction problem: It is addicted to tokens.
Every new generation of agentic AI seems to assume that the answer to complexity is to throw more context at the model, keep longer histories, spawn more calls, loop over more tools, and let the token meter run wild.
The rise of agentic systems, and now projects like OpenClaw, makes that temptation even stronger. Once you give models more autonomy, they do not just consume tokens to answer questions. They consume them to plan, reflect, retry, summarize, call tools, inspect outputs, and keep themselves on track. OpenClaw itself describes the product as an “agent-native” gateway with sessions, memory, tool use, and multi-agent routing across messaging platforms—which tells you exactly where this is going: more autonomy, more orchestration and, unless someone intervenes, a lot more token burn.
That trajectory delights almost everyone selling the infrastructure. If billing is based on tokens, more token consumption looks like growth. If you sell the compute behind those tokens, it looks even better. Google said in its October 2025 earnings call that it was processing more than 1.3 quadrillion monthly tokens across its surfaces, or more than 20 times the volume of a year earlier. Nvidia, for its part, has been leaning hard into the economics of inference and agentic AI, highlighting both the demand surge and the opportunity to sell ever more infrastructure into it.
But companies buying AI should look at this very differently. From the customer’s perspective, explosive token growth is not necessarily a sign of intelligence. In many cases, it is a sign of inefficiency.
More tokens are not the same thing as more intelligence
The current industry narrative often treats token consumption as if it were a proxy for progress. Bigger context windows, more reasoning traces, more agent loops, more memory, more retrieval, more interactions. It all sounds impressive.
Explore Topics
- Artificial Intelligence
- token
Sign in to highlight and annotate this article
Conversation starters
Daily AI Digest
Get the top 5 AI stories delivered to your inbox every morning.
More about
modellanguage modelavailable
Generalist introduces GEN-1 general-purpose model for physical AI
Generalist AI said its latest model marks a significant step toward its mission of creating general intelligence for the physical world. The post Generalist introduces GEN-1 general-purpose model for physical AI appeared first on The Robot Report .
Knowledge Map
Connected Articles — Knowledge Graph
This article is connected to other articles through shared AI topics and tags.
More in Products
Show HN: Wazear – A visual AI orchestrator where agents review each other
Hey folks, For the past month I've been working on a visual AI orchestartor tool that allows users to create a pipeline similar to SDLC. Basically you fire up Wazear, create a project and add your brief. You select the agents (each agent serves a role such as planner, architect, etc...) and set which agent reviews which other agent's work and let it do the work. At any point you can pause the pipeline to review output yourself. You can check it out here: https://wazear.space . Any feedback is welcome. Thank you very much. Best Regards. Comments URL: https://news.ycombinator.com/item?id=47624203 Points: 2 # Comments: 0
Is AI Insurance the Answer to AI Legal Battles?
Chief legal advisors and general counsels might need to consider new AI insurance offerings as part of their strategy to stay on top of their organisation s exposure to AI-related risks, according to Gartner. “As AI incidents surge and insurers increasingly add AI exclusions to traditional policies, companies face growing exposure to legal, financial and regulatory [ ] The post Is AI Insurance the Answer to AI Legal Battles? appeared first on DIGIT .
Akira Hackers Shrink Encryption Timeline to Under One Hour
A notorious ransomware group has been observed leveraging long‑standing exploits and stolen credentials to slip past MFA protections and execute attacks in as little as one hour. Tracking the well-known Akira ransomware group, security researchers from Halcyon witnessed hackers abusing CVE-2024-40766 to gain unauthorised access to SonicWall management interfaces and configuration backups on unpatched devices. [ ] The post Akira Hackers Shrink Encryption Timeline to Under One Hour appeared first on DIGIT .
Discussion
Sign in to join the discussion
No comments yet — be the first to share your thoughts!