Open Menu
AllLocalCommunitiesAbout
lotide
AllLocalCommunitiesAbout
Login

Google DeepMind unveils AlphaEvolve, a Gemini-powered AI coding agent that designs and optimizes advanced algorithms

⁨56⁩ ⁨likes⁩

Submitted ⁨⁨5⁩ ⁨days⁩ ago⁩ by ⁨Pro@programming.dev⁩ to ⁨technology@lemmy.world⁩

https://deepmind.google/discover/blog/alphaevolve-a-gemini-powered-coding-agent-for-designing-advanced-algorithms/

source

Comments

Sort:hotnewtop
  • AbouBenAdhem@lemmy.world ⁨5⁩ ⁨days⁩ ago

    AlphaEvolve verifies, runs and scores the proposed programs using automated evaluation metrics. These metrics provide an objective, quantifiable assessment of each solution’s accuracy and quality.

    Yeah, that’s the way genetic algorithms have worked for decades. Have they figured out a way to turn those evaluation metrics directly into code improvements, or do they just keep doing a bunch of rounds of trial and error?

    source
    • KingRandomGuy@lemmy.world ⁨4⁩ ⁨days⁩ ago

      The general framework for evolutionary methods/genetic algorithms is indeed old but it’s extremely broad. What matters is how you actually mutate the algorithm being run given feedback. In this case, they’re using the same framework as genetic algorithms (iteratively building up solutions by repeatedly modifying an existing attempt after receiving feedback) but they use an LLM for two things:

      1. Overall better sampling (the LLM has better heuristics for figuring out what to fix compared to handwritten techniques), meaning higher efficiency at finding a working solution.

      2. “Open set” mutations: you don’t need to pre-define what changes can be made to the solution. The LLM can generate arbitrary mutations instead. In particular, AlphaEvolve can modify entire codebases as mutations, whereas prior work only modified single functions.

      The “Related Work” (section 5) section of their whitepaper is probably what you’re looking for, see here.

      source
  • PlantJam@lemmy.world ⁨5⁩ ⁨days⁩ ago

    Can’t wait for my boss to come up with half a working solution with this!

    source
  • kaeurennetwo@lemmy.world ⁨5⁩ ⁨days⁩ ago

    What is deep mind?

    source
    • AbouBenAdhem@lemmy.world ⁨5⁩ ⁨days⁩ ago

      It’s just the name of Google’s AI division.

      source
      • entropicdrift@lemmy.sdf.org ⁨5⁩ ⁨days⁩ ago

        It used to be a separate company. They’re the ones who did AlphaZero and AlphaFold too, kind of a big deal at the time.

        source
        • -> View More Comments
  • Williacu@lemmynsfw.com ⁨5⁩ ⁨days⁩ ago

    The announcement of AlphaEvolve represents a major leap forward in the development of autonomous coding agents. Leveraging the Gemini model to iterate, test, and optimize algorithms without direct human input introduces a new class of tools for intelligent algorithm design. This has far-reaching implications not only for computational research but also for applied fields like data science, industrial automation, and custom software development.

    At Kickoff Advertising, where we work with advanced automation to optimize digital marketing campaigns, we’re already exploring the integration of LLM-powered agents for tasks such as automated A/B testing, tracking code optimization, and dynamic content generation based on behavioral data. Tools like AlphaEvolve could further accelerate these processes by delivering real-time, optimized algorithmic solutions.

    We’re particularly interested in learning more about the agent’s underlying architecture—especially its generalization capabilities and the evaluation mechanisms used to assess algorithmic quality.

    source