CommunityNews

CommunityNews

Anthropic CEO: “We Do Not Understand How Our Own AI Creations Work”

Dario Amodei predicts the “MRI for AI” will be here in five to 10 years. And, he outlines three ways to achieve AI model interpretability sooner.

Read in full here:

https://www.techrepublic.com/article/news-anthropic-ceo-ai-interoperability/

Most Liked

jss

jss

I also don’t know why the code I wrote the other day works :squinting_face_with_tongue:

Eiji

Eiji

Company owners does not really need to understand how their projects works under the root as long as they are smart enough to realise it and don’t cause a problem for their employees. That’s said at best they should have at least some ground knowledge about their projects just to not propose a conflicting ideas.

Describing every time why some idea conflicts with another one maid in past only slows down the developer work. If owner is kind and willing to understand it’s fine. However some people expect that developer would blindly implement something as they’re “paying for their work” and therefore they “see no problem”.

If I would be the CEO i.e. a person that many people look what I’m talking about then I would definitely say something that’s completely different:

The LLM is a very complex algorithm and although I don’t understand every aspect of it, I work closely with the developers, because we want to provide the best quality of our services.

Instead of that we’ve got:

People outside the field are often surprised and alarmed to learn that we do not understand how our own AI creations work. They are right to be concerned: this lack of understanding is essentially unprecedented in the history of technology

I sometimes feel that I woke up from an awesome heaven dream and land into a painful reality. Am I the crazy one who believes in “old & good business” dreams or is it another attempt to make people fear about LLM? What’s wrong with the business today? Why they still have funds after they scare the investors? Why it’s considered normal that people in the field takes it as a normal thing?

If I would fund something maybe I would not be interested in many aspects (especially if I would be a big player funding thousands of projects), but I would be definitely interested in what exactly I’m paying for. After such text I would do all I can to take back all the money and make sure that I would never fund any of this company’s project and anything that this CEO was ever touching.

Sure, I understand we don’t live in a world full of trust especially when we are talking about politicians or mainstream medias, but come on … it’s a business! How they can even work with absolutely no trust? In normal world what would you do if an employee would say that they have no idea what they are working on?

Where Next?

Popular General Dev topics Top

First poster: bot
SPWN is a programming language that compiles to Geometry Dash levels. What that means is that you can create levels by using not only the...
New
First poster: bot
Last night I re-read this Steve Yegge article about learning to type as a programmer. I can touch type, but I don’t usually manage to bre...
New
First poster: malloryerik
GitHub - hlissner/doom-emacs: An Emacs framework for the stubborn martian hacker. An Emacs framework for the stubborn martian hacker - G...
New
First poster: bot
GitHub - lucidrains/PaLM-rlhf-pytorch: Implementation of RLHF (Reinforcement Learning with Human Feedback) on top of the PaLM architectur...
New
First poster: dani
The pool of talented C++ developers is running dry. Highly sought after, rarely provided.
New
First poster: peterchancc
Why I like Clojure as a solo developer | Biff. Most of the reasons fall into a few categories: data orientation, the JVM, and the REPL.
New
First poster: bot
Declarative GNOME configuration with NixOS. I adore tinkering with my machine, trying new tools, extensions, themes, and ideas. When I w...
New
CommunityNews
Christian Mills - Testing Intel’s Arc A770 GPU for Deep Learning Pt. 2. This post covers my experience training image classification mod...
New
First poster: dyowee
olmOCR is an open-source tool for converting PDFs to text with high accuracy, preserving reading order and supporting tables, equations, ...
New
CommunityNews
Rendering Action Mailer emails with Phlex components and layouts: Clean, Composable, and Completely Ruby - Blog post by Camillo Visini
New

Other popular topics Top

Devtalk
Hello Devtalk World! Please let us know a little about who you are and where you’re from :nerd_face:
New
PragmaticBookshelf
From finance to artificial intelligence, genetic algorithms are a powerful tool with a wide array of applications. But you don't need an ...
New
AstonJ
Thanks to @foxtrottwist’s and @Tomas’s posts in this thread: Poll: Which code editor do you use? I bought Onivim! :nerd_face: https://on...
New
New
PragmaticBookshelf
Tailwind CSS is an exciting new CSS framework that allows you to design your site by composing simple utility classes to create complex e...
New
rustkas
Intensively researching Erlang books and additional resources on it, I have found that the topic of using Regular Expressions is either c...
New
PragmaticBookshelf
Rails 7 completely redefines what it means to produce fantastic user experiences and provides a way to achieve all the benefits of single...
New
First poster: bot
zig/http.zig at 7cf2cbb33ef34c1d211135f56d30fe23b6cacd42 · ziglang/zig. General-purpose programming language and toolchain for maintaini...
New
New
AstonJ
This is a very quick guide, you just need to: Download LM Studio: https://lmstudio.ai/ Click on search Type DeepSeek, then select the o...
New