Urgent.News

the world's headlines, one feed

Editions

AI

The Tokenpocalypse Is Here: Companies Are Scrambling To Stop Spending So Much on AI

The Tokenpocalypse Is Here: Companies Are Scrambling To Stop Spending So Much on AI There's a fun anecdote from Accenture (apparently via leaked meeting audio recordings) in this 404 Media piece from June 24th: “We’re seeing from some of the data internally at least that it’s actually not our engineers that are driving the token consumption. It’s a lot of the non-engineers that are doing some of…

Abstract editorial illustration

The "Tokenpocalypse" has arrived, with companies scrambling to curb their expenditure on AI. A recent report from 404 Media reveals that non-engineers are consuming a significant amount of AI tokens, challenging the notion that only engineers are responsible for this surge. Accenture's AI strategy lead, Justice Kwak, shared internal data showing that PDFs are a major culprit in token consumption.

This revelation prompted Stuart Henderson, Accenture's client group lead, to question whether this was indeed the case, suggesting that converting PDFs into markdown files may be a contributing factor. Kwak confirmed that this is what the company's own data indicates. In a bid to spread awareness, Kwak humorously noted that if Accenture can figure out the adverse effects of PDFs, they might be able to disseminate this message to the broader business community.

Written by urgent.news from Simon Willison's reporting — not their text. Machine-written — it may contain errors, so check the original before relying on it.

Read the original at simonwillison.net →

More in AI

Can you run a GNN without anyone seeing the data? Journal of our experiments on privacy-preserving GNN inference on microcontrollers

I have been keenly interested in how to make Graph Neural Networks (GNN) run securely on resource-constrained devices. The question driving this work is deceptively simple: can three microcontrollers…

  • Researchers demonstrate privacy-preserving GNN inference on microcontrollers
  • Graph topology public, only node features and model weights protected
  • Secure matrix multiplication handled via Replicated Secret Sharing