Wildroot

Queries may use an external AI service. Details

Wildroot Browser

TodayAISimilar posts

Similar posts

“The economist Sendhil Mullainathan, in writing about the dangers of AI, cited the long-tail phenomenon (which I described in chapter 6) in his notion of 'tail risk’:” «We should be afraid. Not of intelligent machines. But of machines making decisions that they do not have the intelligence to make. I am far more afraid of machine stupidity than of machine intelligence. Machine stupidity creates a tail risk. Machines can make many many good decisions and then one day fail spectacularly on a tail event that did not appear in their training data. This is the difference between specific and general intelligence.»

Show more of this post

“Or as the AT researcher Pedro Domingos so memorably put it, 'People worry that computers will get too smart and take over the world, but the real problem is that they're too stupid and they've already taken over the world.’” #MelanieMitchell wrote this in 2019. It’s 2026 and I have not seen any improvement in this regard: no advancement in building systems that understand rather than mimic, and hence remain fragile to the long tail of events absent from the training data. #AI #LLMs

Opinion

3 posts close in meaning, closest first

  1. I was sharing with my wife my (current) preferred approach to AI produced code by engineers. It's borrowed from her PhD - thesis defence. If an engineer can defend their code in a thesis defence, I have no concern about how it was produced. What is thesis defence for software changes? They should be able to explain, without using an llm: * What is the objective for the code change * Why did they choose this approach * What other approaches they considered * What unrelated/unexpected things they found along the way * What are the limitations of chosen approach * What is the wider/longer term impact of choosing this approach

    Show more of this post

    * Who else did they collaborate with, or who else understands their chosen approach to be able to answer these questions if needed Notice I don't want them to explain their work. I can read the code and understand it. I want to understand how well researched this approach is, and how well they understand it. I also want to ensure that they have put in at least as much effort (to produce the code or to prepare for the theses defence) as the reviewer will put in. #ai #softwareengineering

    Similar topic and wordingOpinionAI
  2. And I'm not coming at this from a "I'll never use AI, it's evil" stance. I use AI at work a fair bit. But I use it in a way that I still own, deeply understand, and am able to explain and modify anything that I produce - whether I built it with or without AI. Around me I see work being pushed where the author can't answer a few questions about it, it is clearly beyond their usual ability, and has no chance of being maintainable workout AI. This is just lazy folks using AI lazily. I can handle this (at a cost). They don't understand it but I can, and I can make it simpler and maintainable. What really has me reconsidering my approach is the Software Factory approach I am seeing being adopted by some of the smarter people I know. They are actively making a choice that they don't need to know how the system being produced works in details. As long as it produces desired behaviour/output, it is all that matters. #ai #softwareengineering

    Similar topic and wordingOpinionAI
  3. @LabSpokane@mas.to

    Fucking exactly. LLMs only “work” because they literally reduced the probability of the next token in order to get interesting outputs. Not more correct outputs, but they got the models to babble on longer without repeating themselves. How the fuck can anyone validate any of this shit for consistency with effectively a random number generator in-the-loop? Why are we still here talking about this? For several trillion dollars surely we could have made more deterministic tools to solve our problems that didn’t require hundreds of square miles of compute and dozens dozens of nuclear reactors’ worth of power. RE: floss.social/@janriemer/110204…

    Similar topic and wordingOpinionAI