Ola Reviews

Tag: Meta

  • Facebook’s New “Your Algorithm” Tool: How to Take Control of Your Feed

    Facebook’s New “Your Algorithm” Tool: How to Take Control of Your Feed

    For years, Facebook users have complained that their feed feels random. Now Meta is offering a direct fix. A new tool called Your algorithm lets you tell Facebook, in plain words, what you want to see more or less of. If you have ever wanted to control your Facebook feed instead of fighting it, this is the biggest change in a long time. Here is what it does and how to use it well.

    What “Your Algorithm” Actually Does

    Announced on 13 October 2026 in a Meta newsroom post, the feature gives you a dedicated space to add and manage the topics you want in your Feed and Reels. You can type requests in your own words, such as more home cooking videos or less celebrity gossip. Meta says your recommendations respond in real time, even within the same session.

    Behind the scenes, Meta is now using large language models in its ranking systems. That is the technology that lets Facebook understand a written request rather than relying only on what you click and watch.

    How to Use It, Step by Step

    The feature is rolling out gradually, so menus may look slightly different on your account. The basic steps are:

    • Tap the three-dot menu on any post in Feed or Reels.
    • Choose Your algorithm to view and edit the topics shaping your recommendations.
    • Add topics you want more of, and remove or reduce ones you are tired of.
    • On individual posts, tap Interested or Not interested.
    • Answer any follow-up questions. Facebook may ask why you liked or disliked something, and a sentence of detail helps.

    Be specific. A request like “local footy highlights and Aussie travel tips” gives the system far more to work with than “sport”.

    Is It Available in Australia?

    Meta says the feature is rolling out but has not named countries, languages or a completion date. Its track record offers a clue. Instagram introduced its own Your Algorithm for Reels in the US in December 2025, then expanded it across Feed, Reels and Explore globally in English in June 2026, as TechCrunch noted. Meta’s similar tool on Threads, Dear Algo, is already available in Australia. If you do not see it yet, update the Facebook app and check again over the coming weeks.

    The Privacy Trade-Off

    Telling an algorithm exactly what you want is useful, but it also hands Meta clearer information about your interests. That information can shape ads as well as posts. A few habits help:

    • Avoid typing sensitive details, such as health conditions or money problems, into topic requests.
    • Review your topics now and then, and delete ones you no longer want linked to your account.
    • Check your ad preferences in Facebook’s settings alongside your feed topics.

    Our guide to [Link: social media privacy settings in Australia] walks through the settings worth changing.

    What It Means for Creators and Businesses

    For pages and creators, this adds a new signal they cannot see. If many users ask for less of a certain style of post, reach may drop without an obvious reason. The safest response is to post genuinely useful content that people would ask for by name. Scheduling tools still matter, as our comparison of Buffer vs Publer explains, but relevance now matters even more.

    A Feed That Finally Listens

    Your algorithm turns Facebook’s ranking system from something that watches you into something you can talk to. Used carefully, it can make scrolling more useful and less frustrating. The best results will come from clear, specific requests and a regular check on what you have shared.

    Frequently Asked Questions

    Do my topic changes last forever?

    Meta has not said whether added topics are permanent or fade over time. Treat them as something to revisit every few months.

    Does this remove ads?

    No. Meta describes Your algorithm as a way to shape recommended posts and Reels. Ads are managed through Facebook’s separate ad preferences, so check those too.

    Will posts from friends disappear?

    Meta has not said the tool hides posts from friends. It is designed to shape recommendations, and you can still manage friends and pages you follow through Facebook’s existing feed preferences.

    Can I undo a change?

    Yes. You can remove topics from Your algorithm at any time, and your recommendations will adjust.

  • Meta’s Muse Wonder Explained: What Its First Frontier AI Model Means for Everyday Users

    Meta’s Muse Wonder Explained: What Its First Frontier AI Model Means for Everyday Users

    Meta has spent years chasing the leaders in the AI race. With Muse Wonder, it is trying to join them. Released on 12 October 2026, it is the company’s first true frontier model, meaning it aims to match the most capable systems from rivals such as OpenAI and Anthropic. Mark Zuckerberg’s launch post passed 4 million views on X within a day, a sign of how closely the tech world is watching. So what does this new Meta AI model actually do, and should everyday Australians pay attention?

    Two Models, Two Very Different Doors

    Muse Wonder is really a pair of models. Meta’s Superintelligence Labs released Muse Wonder and Muse Wonder Full together as the company’s first attempt at a frontier model.

    The standard Muse Wonder is the one most people can reach. It is available through the Meta Model API and Muse Code, Meta’s coding tool. The larger Muse Wonder Full is more restricted. It is only offered to Muse Code subscribers and is not sold through the API, and both models also power the Muse agent, as TechCrunch reported at launch.

    This split matters more than it first appears. Many of the strongest results Meta promoted come from the Full version, so the model you can actually pay for by the token is not always the one topping the charts.

    Memory Is the Big Selling Point

    Most chatbots forget you once a conversation ends, or keep only a short list of saved facts. Meta is betting on active memory, which means retaining and connecting what you share over long periods. On Meta’s own testing, Muse Wonder scored higher than rival models on the MemoryBench benchmark, which measures this kind of long-term recall.

    For everyday users, this could mean an assistant that recalls your diet, your work projects or your travel plans without reminders. It also raises a fair privacy question. An assistant that remembers more also stores more, so it is worth checking what is kept and how to delete it. Our guide to AI privacy rules in Australia explains your rights.

    How it compares on independent tests

    Benchmarks run by the company that built a model deserve some caution. Meta says Wonder and Wonder Full achieve scores comparable to Fable 5.1 and GPT-6 Astra. Independent testing paints a more modest picture. Artificial Analysis scored Muse Wonder 52 on its Intelligence Index, compared with 56 for Wonder Full and 58 for Claude Opus 5.5. In short, it is competitive, but not a clear winner.

    What It Costs, and What That Signals

    Meta’s earlier models were known as budget options. That era appears to be over. Through the Meta Model API, Wonder costs US$4 per million input tokens and US$20 per million output tokens, roughly the same as GPT-6 Sol or Opus 5.5. A token is a small chunk of text, so these prices mostly matter to businesses building apps.

    Individuals are more likely to pay through a subscription. Wonder Full has no API price and comes with Muse Code’s Power plan at US$50 a month or its Maximum plan at US$200 a month, according to explainx’s breakdown of Meta’s pricing pages. Prices are listed in US dollars, so Australian buyers should expect the final cost to move with the exchange rate. You can compare it with rival tools in our test of Reasoning vs Rock-Bottom Pricing: Opus 5.5 and GPT-6 Luna.

    A Rocky First Day

    No big launch goes perfectly. Within hours, users complained that Muse Wonder was eating through their usage limits far faster than expected. Alexandr Wang, who leads Meta’s AI work, posted that the team had found and fixed the problem. Quick fixes are reassuring, but they also show this is still very new software.

    Should Australians Try It?

    For developers and businesses, Muse Wonder is worth testing, especially for work that depends on long-term memory, such as customer support or personal assistants. For casual users, there is less urgency. If you already pay for an AI assistant, compare features and the real cost in Australian dollars before switching. Our roundup of best AI assistants in Australia can help you weigh the options.

    A Serious Contender, Not a Knockout

    Muse Wonder marks the moment Meta stopped competing mainly on price and started competing on capability. Its focus on memory could make AI assistants feel far more personal, though that same strength calls for careful attention to privacy. For now, it is a strong new option rather than a clear champion, and the smartest move is to test it on the tasks you actually care about.

    Frequently Asked Questions

    Is Muse Wonder free to use?

    No. Meta charges per use through its API, and the more powerful Full version requires a paid Muse Code plan. Check current plan prices before you sign up, since launch pricing can change.

    Is Muse Wonder open source?

    No. Neither Meta’s launch post nor its model card mentions releasing the model weights. That is a shift from Meta’s earlier open Llama models.

    What is Muse Wonder best at?

    It is built for agents that work on your behalf for hours, rather than for quick one-off answers. Coding, long-running tasks and remembering context are its main strengths.

    Can it read images and audio?

    Yes. It accepts text, images, video and audio as input, and it replies in text.