• 280: Evidently, The Cloud Pod Was Always Right

  • Oct 31 2024
  • Length: 56 mins
  • Podcast

280: Evidently, The Cloud Pod Was Always Right

  • Summary

  • Welcome to episode 280 of The Cloud Pod, where the forecast is always cloudy! This week Justin, Jonathan, Ryan, and Matthew are your hosts as we travel through the latest in cloud news. This week we’re talking more about nuclear power, some additional major employee shakeups, Claude releases, plus saying RIP to CloudWatch Evidently and hello to Azure Cobalt VMs.

    Titles we almost went with this week:
    • The cloud providers are colluding on Nuclear Power
    • I fear our AWS AI nightmare might get worse without Dr. Matt Wood.
    • I’m a glow with excitement about nuclear cloud power
    • Plainly no one else knew what “CloudWatch Evidently” did either
    • We sing a Claude Sonnet about Nuclear Power
    • Evidently, The Cloud Pod was always right
    • Amazon goes nuclear while their AI VP goes AWOL
    A big thanks to this week’s sponsor: We’re sponsorless! Want to get your brand, company, or service in front of a very enthusiastic group of cloud news seekers? You’ve come to the right place! Send us an email or hit us up on our slack channel for more info. AI Is Going Great – Or How ML Makes All It’s Money

    00:53 Introducing computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku

    • Anthropic is announcing the upgraded Claude 3.5 Sonnet and a new Model Claude 3.5 Haiku.
    • Claude 3.5 Sonnet delivers across the board improvements over its predecessor, with particularly significant gains in coding — an area where it already leads the field (per anthropic).
    • Claude 3.5 Haiku interestingly matches the performance of Claude 3 Opus, the prior largest model, on many evaluations at the same cost and similar speed to the previous generation of Haiku.
    • Claude 3.5 Sonnet also includes a groundbreaking new capability in beta: Computer Use.
    • Available today as an API, developers can direct Claude to use computers the way people do – by looking at a screen, moving a cursor, clicking buttons and typing text.
    • Claude 3.5 is the first frontier AI model to offer this capability.
    • Anthropic warns the feature is still experimental – at times cumbersome and error-prone. As well as things that are effortless for a human are still difficult including scrolling, dragging or zooming.
    • The idea is to make Claude complete individual tasks, without always needing to leverage an API, like clicking in a GUI, or uploading a file from a computer. These types of solutions are typically found in Build and Test like scenarios with tools such as Saucelabs or Browserstack.
    • To do this, Claude was built to perceive and interact with computer interfaces. You can use data from my computer to fill out this online form or check a spreadsheet, move the cursor to a web browser, navigate to the relevant web pages, select the data for the spreadsheet and so on.

    3:06 Jonathan – “If you can take pictures of the screen, then it can identify where buttons and things are without having to know the name of the objects in the DOM and stuff like that. So you could say, give me instructions, click on this, click on this, click on this, do this stuff. It would be really easy to automate tests that way instead of having to know the names of the divs and things on a page, especially for web testing. Because if a developer changes those, the...

    Show More Show Less
activate_Holiday_promo_in_buybox_DT_T2

What listeners say about 280: Evidently, The Cloud Pod Was Always Right

Average customer ratings

Reviews - Please select the tabs below to change the source of reviews.