TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Artificial Intelligence Isn't Ready for Mass Application

2 pointsby robdoherty25 months ago

1 comment

gibbitz5 months ago
Why are we relying on hardware to fix this problem? Software efficiency has been in the decline for decades and each layer is built on the shitty layer that came before it. The architecture of LLMs is like brute forcing problem resolution so no wonder it&#x27;s inefficient. Imagine if the way we solved multiplication in calculators was by trying every number until the right one showed up. Somehow this is the LLM calculator we are marketing right now. This is not how we get to AGI, unless we use a dollar amount to determine when we reach it and not cognition.<p>We are quickly approaching the quality ceiling afforded by serving Capitalists who MVP every product just to the point they can make the most profit and stop. Advancement for quality with no return on investment is shot down without question. If the middle ages served the market and not the church, all the cathedrals would have been built of wood and would be gone now.<p>Articles like this are just managing the consumer&#x27;s expectations so the companies can continue to under deliver and avoid costly R&amp;D. Why are we tolerating this?