All the big tech companies are trying to outdo each other with “Hey, look what our artificial intelligence can do now” updates, but on Thursday Facebook’s parent company Meta announced its latest update with a version called Llama 3. An update has been published.
Unlike previous Llama releases, Llama 3 has a dedicated website (meta.ai) where you can use the tool for free. This is an official riff on ChatGPT. There is no charge. Go and talk to them and see what they can do for you. The problem, of course, is logging in with your Facebook account.
“The big day is here. We’re releasing a new version of Meta AI, the assistant you can ask any question you want across apps and glasses,” Metaplatform CEO Mark Zuckerberg told USA TODAY. said in a statement sent. “We’ve built this feature into the search box at the top of WhatsApp, Instagram, Facebook, and Messenger, so whenever you have a question, you can ask it there. We’re also releasing a bunch of unique authoring features. Meta AI generates and updates images in real-time as you type to create animations and create high-quality images very quickly.
Some of my friends are already seeing the “Ask Meta AI anything prompt” at the top of their search feeds, but mine isn’t showing it yet. However, meta upgrades are typically deployed this way. Some people get it right away, while others have to wait a little while.
Will it change your life in any way? Absolutely not. Do you care? perhaps. Let’s dive in.
What does meta AI do?
The most interesting thing I’ve seen so far is generating realistic animated characters and images that update while you type prompts.
Just enter what you want to see and the image will start generating as you type. For example. Want to see a curly-haired girl flying dragonflies? Type in “curly girl” and you’ll see it there. As you can see in the example above, the image continues to adapt to the story as you type. The only limit is your imagination. It’s like you can choose your own adventure, but there are no dire consequences.
The great thing about real-time image generation is that you don’t have to wait for the entire prompt to be written. Instead, it starts “seeing” each object as the user writes about it, and adjusts the final image in real time. This requires powerful hardware and software, but it’s interesting to see it in action.
What could go wrong?
This is a huge waste of time and is clearly a very powerful tool. However, this is an issue plaguing similar tools such as Microsoft and Midjourney, so you should check whether Meta’s obscene content filter is fully functional.
As for guardrails, most popular AI image generators, such as Microsoft, have blacklisted keywords that can generate images of celebrities such as movie stars or prominent politicians. The datasets these models use typically don’t include “ordinary people,” so if you feed your friends’ names into the AI, it won’t generate anything similar to them.
However, there are always ways around these guardrails. An avid user may find a way to feature a particular celebrity in an image or use the system to create his NSFW (Not Safe for Work) content. This is something every other AI image generator has been fighting against since day one.
Meta AI to find friends and win arguments
What about generating fake images that people think are real, like politicians in the next election? I’ll get back to that in a moment, but first, let’s talk about the company’s new Meta AI Here are some other things you can do.
◾ Acts as a personal assistant. “We believe Meta AI is the most intelligent AI assistant available for free, and now available in more countries through our app, it can help you make dinner based on what’s in your fridge. It helps you plan and study for tests,” the company said in a statement. I see this as a way to win arguments with people on the app and in real life.
◾ Enable more powerful searches. According to the company, “Using Meta AI in feed, chat, search, and more across our apps, you can get work done and access real-time information without leaving the app you’re using. “apparently.
◾ Create faster and more customized images. “Create album album artwork for your band, apartment decor inspiration, animated custom GIFs, and more.”
During a video call demonstration earlier this week, a Facebook spokesperson said they expect people to use MetaAI “for fun and to be more social.”
To me, that means they’re trying to make Facebook more fun for more than just our parents, and make Instagram more interesting so that young people don’t keep abandoning Facebook for TikTok. (Facebook still has about 3 billion monthly active users, or about a third of the world’s population, according to digital marketing site Social Surge. That’s a harsh statement, though.) , which is true in a sense.)
Rama 3 will be available in 12 countries including Canada, Australia, and South Africa, all of which will introduce Meta’s AI for the first time.
Is meta AI safe?
Llama 3 is Meta’s next generation open source large-scale language model. This is similar to ChatGPT which is about to announce an upgraded model GPT-5. Such models power AI-based apps such as chatbots, image generators, and virtual assistants.
Upgrading from Llama 2 (released July 2023) to Llama 3 should This means more accurate responses, more human-like conversations, and a broader knowledge base for the AI apps that use it. Meta is not the first to debut a super-powerful artificial intelligence model. Of course, that honor went to her OpenAI with ChatGPT. But Meta is working hard to catch up. Large language models such as ChatGPT and Llama are very versatile and can be used for a variety of purposes. These are the brains behind AI chatbots, developer coding tools, automated writing assistants, and many other tools that are becoming more and more popular every day.
Building large language models (Proprietary LLM) means we can power the next generation of extremely powerful and user-friendly apps, and that’s clearly something Meta is interested in. Meta announced his Llama 2 last July, and its successor is already on the way. Find out everything you need to know about how many resources Meta is putting into the project.
Llama 3’s power comes from Meta’s even larger data base of AI. The company says Llama 3 draws from the vast pool of information generated by the people who use its products. However, we promise that no private or personal information will be input into its language model.
That’s a heartwarming promise, but can we guarantee that sensitive information won’t find its way into these AI datasets? Unless history tells us otherwise, that’s the cost of using social media.
Can a chatbot using Llama 3 technology be used for fraudulent purposes? Can it impersonate you or someone you know? Can it effectively “deepfake” your online identity? , could a scammer tarnish your name or worse?
Mehta says that such a scenario is impossible thanks to various safeguards, but when we joined Facebook more than a decade ago, there was a possibility that one day our content would feed the AI. I could never have imagined that it would happen. Wrestling with these realities may be uncomfortable for many longtime Facebook users.
Specifically, Meta says it does not use any “public figures” to generate images or videos. When you type in someone’s name, a computer can generate what it thinks the person with that name looks like, which isn’t necessarily accurate.
It also has some NSFW safeguards built in to protect against pornography, racism, and other issues that plague generative AI, but we don’t know how well those guardrails will work.
Why is Meta investing so much in AI?
Meta does this because large language models like Llama 3 allow you to add new features to your app, like adding AI to search functionality on Facebook and Instagram. This adds a layer of smarts that reduces friction between users and the apps they prefer.
What’s going on with the Facebook app?:Why does the Facebook app look different? Meta rolls out new full-screen video player
By building its own high-performance AI, Meta can also sell licenses to companies who want to use it in their projects. ChatGPT runs on the same licensing business model, granting businesses and their users access to AI by paying a specific fee. Meta makes a lot of money using his LLM in a variety of ways, including advertising (specifically, think auto-generated ads that interest users) and selling access to this AI. I’m trying.
Facebook has one of the largest accumulations of corporate data thanks to more than a decade of user-generated content in the form of posts, comments, replies, and more. This data trains the llama how humans speak and make connections between words and objects, places, and concepts. The smarter you are, the more convenient you should be.
That’s the theory, but I’ve covered technology long enough to be skeptical at best.
