Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku

anthropic.com

Post reply on HN
Print view View original

691–700 of 758 posts

Re: Computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku

#691

Post by bonoboTP »

I've been saying this is coming for a long time, but my really smart SWE friend who is nevertheless not in the AI/ML space dismissed it as a stupid roundabout way of doing things. That software should just talk via APIs. No matter how much I argued regarding legacy software/websites and how much functionality is really only available through GUI, it seems some people are really put off by this type of approach. To me, who is more embedded in the AI, computer vision, robotics world, the fuzziness of day-to-day life is more apparent.

Just as how expert systems didn't take off and tagging every website for the Semantic Web didn't happen either, we have to accept that the real world of humans is messy and unstructured.

I still advocate making new things more structured. A car on wheels on flattened ground will always be more efficient than skipping the landscaping part and just riding quadruped robots through the forest on uneven terrain. We should develop better information infrastructure but the long tail of existing use cases will require automation that can deal with unstructured mess too.

Re: Computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku

#692

Post by marvinkennis »

speleding wrote: post #616
LASR wrote: post #344

This is actually a huge deal. As someone building AI SaaS products, I used to have the position that directly integrating with APIs is going to get us most of the way there in terms of complete AI automation. I wanted to take at stab at this problem and started researching some daily busineses and how they use software. My brother-in-law (who is a doctor) showed me the bespoke software they use in his practice. Runni…

I'm a bit skeptical about this working well enough to handle exceptions as soon as something out of the ordinary occurs. But it seems this could work great for automated testing. Has anyone tried asking "use computer" to do "Please write a selenium/capybara/whatever test for filling out this form and sending it?" That would take away some serious drudge work. And it's not a big problem if it fails, contrary to when i…

Momentic (W24) is doing this! No affiliation, but they've made some solid progress https://momentic.ai/

Re: Computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku

#693

Post by ttyprintk »

amelius wrote: post #639
LASR wrote: post #344

This is actually a huge deal. As someone building AI SaaS products, I used to have the position that directly integrating with APIs is going to get us most of the way there in terms of complete AI automation. I wanted to take at stab at this problem and started researching some daily busineses and how they use software. My brother-in-law (who is a doctor) showed me the bespoke software they use in his practice. Runni…

I think it was pretty clear from the beginning that the whole AI thing is going to be winner-takes-all. If you're in some niche doing AI development work, you are going to be outcompeted by more generalized AI at some point.

With one big exception: your general AI might dominate the business case, but my specialized one can craft raw packets. I’m the one who names the price, not you.

Re: Computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku

#694

Post by a9dhalaan »

bloedsinnig wrote: post #46
diggan wrote: post #9

I still feel like the difference between Sonnet and Opus is a bit unclear. Somewhere on Anthropic's website it says that Opus is the most advanced, but on other parts it says Sonnet is the most advanced and also the fastest. The UI doesn't make the distinction clear either. Then on Perplexity, Perplexity says that Opus is the most advanced, compared to Sonnet. And finally, in the table in the blogpost, Opus isn't eve…

Big models / huge models take weeks / month longer than the smaller ones. Thats why they release them with that skew

I don't think that's quite it. They had it on their website before this, that opus 3.5 was coming soon, now they've removed that from the webpage.

Also, Gemini ultra 1.0, was released like 8 months ago, 1.5 pro released soon after, with this wording "The first Gemini 1.5 model we’re releasing for early testing is Gemini 1.5 Pro"

Still no ultra 1.5, despite many mid and small sized models being released in that time frame. This isn't just an issue of "the training time takes longer", or a "skew" to release dates. There's a better theory to explain why all SoTA LLM companies have not released a heavy model in many months.

Re: Computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku

#695

Post by eitally »

aduffy wrote: post #374
LASR wrote: post #344

This is actually a huge deal. As someone building AI SaaS products, I used to have the position that directly integrating with APIs is going to get us most of the way there in terms of complete AI automation. I wanted to take at stab at this problem and started researching some daily busineses and how they use software. My brother-in-law (who is a doctor) showed me the bespoke software they use in his practice. Runni…

This has existed for a long time, it's called "RPA" or Robotic Process Automation. The biggest incumbent in this space is UiPath, but there are a host of startups and large companies alike that are tackling it. Most of the things that RPA is used for can be easily scripted, e.g. download a form from one website, open up Adobe. There are a lot of startups that are trying to build agentic versions of RPA, I'm glad to s…

My impression is that actually solving this classic RPA problem with AI is exactly the raison d'etre of AI21Labs with their task specific models[1]. They don't have the biggest or best general purpose LLM, but they have an excellent model that's been pre-trained on specific types of business data and also made available for developers using simple APIs & "RPA-style" interfaces.

[1] https://www.ai21.com/use-cases

Re: Computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku

#696

Post by a9dhalaan »

thenameless7741 wrote: post #583
simonw wrote: post #499

Claude 3.5 Opus is no longer mentioned at all on https://docs.anthropic.com/en/docs/about-claude/models Internet Archive confirms that on the 8th of October that page listed 3.5 Opus as coming "Later this year" https://web.archive.org/web/20241008222204/https://docs.anth... The fact that it's no longer listed suggests that its release has at least been delayed for an unpredictable amount of time, or maybe even cancel…

Before anyone reads too much into this, here's what an Anthropic staff said on Discord: > i don't write the docs, no clue > afaik opus plan same as its ever been

Maybe he's not high level enough employee to have any say in the product roadmap, and he's behind on leadership planning?

Re: Computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku

#697

Post by DebtDeflation »

bonoboTP wrote: post #691

I've been saying this is coming for a long time, but my really smart SWE friend who is nevertheless not in the AI/ML space dismissed it as a stupid roundabout way of doing things. That software should just talk via APIs. No matter how much I argued regarding legacy software/websites and how much functionality is really only available through GUI, it seems some people are really put off by this type of approach. To me…

>it seems some people are really put off by this type of approach

As someone who has had to interact with legacy enterprise systems via RPA (screen scraping and keystroke recording) it is absolutely awful, incredibly brittle, and unmaintainable once you get past a certain level of complexity. Even when it works, performance at scale is terrible.

Re: Computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku

#700

Post by idopmstuff »

bonoboTP wrote: post #691

I've been saying this is coming for a long time, but my really smart SWE friend who is nevertheless not in the AI/ML space dismissed it as a stupid roundabout way of doing things. That software should just talk via APIs. No matter how much I argued regarding legacy software/websites and how much functionality is really only available through GUI, it seems some people are really put off by this type of approach. To me…

It's basically the digital equivalent of humanoid robots - people object because having computers interact with a browser, like building a robot in the form of a human, is incredibly inefficient in theory or if you're designing a system from scratch.

The problem is that we're not starting from scratch - we have a web engineered for browser use and a world engineered for humanoid use. That means an agent that can use a browser, while less efficient than an agent using APIs at any particular task, is vastly more useful because it can complete a much greater breadth of tasks. Same thing with humanoid robots - not as efficient at cleaning the floor as my purpose-built Roomba, but vastly more useful because the breadth of tasks it can accomplish means it can be doing productive things most of the time, as opposed to my Roomba, which is not in use 99% of the time.

I do think that once AI agents become common, the web will increasingly be designed for their use and will move away from the browser, but that probably take a comparable amount of time as it did for the mobile web to emerge after the iPhone came out. (Actually that's probably not true - it'll take less time because AI will be doing the work instead of humans.)

Post reply on HN