AI Insider No. 19

(Michelle Johnson via Midjourney)

It’s OK to Take a Break from AI News

By Michelle Johnson, AI Insider

Are you feeling overwhelmed and left behind by all of this AI stuff?

As much as I love it, I, too, feel overwhelmed sometimes. There are quiet periods, but then the floodgates open again, ushering in a slew of new announcements.

It’s happening right now. This past week has been an avalanche of AI announcements.

Sometimes, the inclination is to check out. And that’s OK. Give yourself permission to ignore it all. And don’t beat yourself up because you’re not using every new AI geegaw that pops up.

You don’t need to learn everything at once. Some of it will come to you when it’s time.

Maybe it will show up with changes at work or in an app you use daily. Maybe your kids will show you something cool. Maybe you’ll click a link in this newsletter.

You’ll get there when you get there. Or not. It’s ok to just be an observer at this point. As I keep saying, it’s all currently a work in progress. Everything is a beta. We know it will change tomorrow.

In the meantime, you be you and take a break if you need to. As for me, I’ll continue to keep an eye on what’s coming, but maybe not cram in so much every day.

Exhale. Peace, everyone.

Browse with Bing appears in the settings for ChatGPT Plus. (Screenshot)

ChatGPT is Browsing Again!

I woke up Thursday morning to the news that ChatGPT is browsing the web again via a beta feature called “Browse with Bing.” Yay!

You may recall that OpenAI pulled the plug on ChatGPT’s connection to the internet a couple of months ago. They took it in for an overhaul to stop it from doing things like getting behind paywalls for free.

Unfortunately, no Internet means no access to current events. That made ChatGPT less useful for me.

Browse with Bing is currently available for paid Plus and Enterprise accounts. I’ve read that it will be coming to the free version soon.

[Update: After using it for a few days, I’m less than impressed with Browse with Bing. I think OpenAI may have locked things down a bit too much. Browse with Bing returns very brief, sparsely detailed responses, which Bing tends to do. That’s not very ChatGPT-like. I don’t like having to ask it to expand on this or that. We’ll see how this shakes out going forward.] -— Michelle Johnson

Embedded cameras in the Meta Ray-Ban smart glasses allow users to take photos and stream video. (Meta)

Meta (aka Facebook) Goes All Out with AI

By Bing and Bard, for AI Insider

Thursday’s Meta Connect event was all about AI, with the company announcing a number of new products and features that leverage this powerful technology. Here’s a rundown of the AI-related announcements:

Meta Ray-Ban Smart Glasses

One of the most talked-about announcements was the collaboration between Meta and Ray-Ban to create smart glasses. These aren’t your average pair of shades! The glasses come with built-in cameras, open-ear audio, and social sharing. You can also livestream without pulling out your phone! They’re designed to integrate AI technology into our everyday lives in a stylish and practical way. Imagine being able to capture your favorite moments, listen to your favorite tunes, or even access real-time information — all from your glasses!

AI Features in Messenger, WhatsApp, and Instagram

Meta also announced a host of new AI features across its messaging platforms — Messenger, WhatsApp, and Instagram.

In Messenger, you’ll soon be able to chat with an AI assistant. This assistant can generate answers to your queries and even generate images. It’s like having a personal assistant right in your chatbox!

WhatsApp is also getting an upgrade with the introduction of AI stickers. These stickers are generated by AI based on your text prompts. So you can create unique stickers for your chats in seconds.

Instagram is getting two new AI-powered image editing features: Restyle and Backdrop. With Restyle, you can ask the AI to add an effect to your image. For example, you can make an image look like it was painted with watercolor. Backdrop allows you to change the background of your image based on a description you provide. [Ed. note: Why do we care? Well, instead of using pre-built filters to change images, you can create anything that you can think of by adding a customized prompt. You could remove a boring background and replace it with a beach and palm trees.]

Meta also introduced its new image generator called Emu. This tool will power features like AI stickers and image editing across Meta’s apps. You can create custom stickers that represent a thought or idea that’s just right for your chat.

Meta Quest 3

Meta’s latest VR headset, the Quest 3, is powered by a new AI chip that promises to deliver significant performance improvements. This will enable developers to create more immersive and realistic VR experiences.

Meta is also developing new AI characters and friends that users can interact with in the metaverse. These characters will be able to learn users’ preferences and interests over time and provide them with more personalized experiences.

Even if you’re not a techie, you can expect a few things from Meta’s AI announcements. If you’re into Meta’s VR experience via the Quest headset or use apps like Instagram, What’s App, and Messenger, these features will come right to you. No need to download or search for them.

Overall, the Meta Connect event showed that AI is a top priority for the company. By investing in AI, Meta aims to keep innovating and enhancing the user experience


Amazon Invests Billions in Anthropic, Maker of Claude AI

By Bing, for AI Insider Inc. announced a major investment in Anthropic, a startup that specializes in generative artificial intelligence. The deal, which could reach up to $4 billion, will give Amazon access to Anthropic’s powerful AI models and help Anthropic scale up its operations.

Anthropic is the company behind chatbot Claude, which is designed to be more secure than other chatbots on the market.

Amazon, the world’s largest provider of cloud computing services, has been lagging behind its rivals in the field of generative AI. Microsoft Corp., for instance, partnered with OpenAI, the creator of ChatGPT, the popular chatbot.

By investing in Anthropic, Amazon hopes to catch up with the competition and become a major player in generative AI. As part of the deal, Anthropic will move most of its software to Amazon Web Services (AWS) data centers and use Amazon’s homegrown chips to train its models. 

The investment in Anthropic is not only a strategic move for Amazon but also a vote of confidence in the startup. Anthropic was founded by former OpenAI researchers and engineers who wanted to create more ethical and transparent AI systems.

Anthropic plans to use Amazon’s investment to cover the substantial costs of training and operating large-scale AI models. The startup also says it will maintain its independence and vision and continue to pursue its goal of creating beneficial AI for humanity.

Snap a picture with the ChatGPT app, then ask away! (OpenAI)

ChatGPT Gets ‘Eyes and Ears’

[Ed. Note: If you see people gushing about how ground-breaking this is, pls note the following. Bing and Bard have had a similar feature for months. And Midjourey, the image generator, can “/Describe” an image after you upload it. I’m not sure why there’s so much whooping about this announcement. Maybe I’m missing something. It’s not exactly new or revolutionary. The big news for me is that it’s baked into ChatGPT. ]

By Perplexity and Michelle Johnson, for AI Insider

OpenAI has announced a significant update to its popular AI assistant, ChatGPT, by adding voice and image-based capabilities to the platform. The update allows users to have voice conversations with the chatbot, as well as upload images and ask ChatGPT questions about the content.

This puts ChatGPT into the “multimodal” category, meaning it can respond to text, voice, and images. However, the jury is out on how well these features work. Stay tuned.

In the meantime, the updates allow users to do things like verbally ask ChatGPT to create a bedtime story on the spot, use vocal prompts to guide the narrative, or simply ask a question and receive a spoken-word response.

You can purportedly take pictures of objects or situations and ask the chatbot for assistance, such as troubleshooting why a grill won’t start, using the contents of a fridge to plan a meal, or analyzing a complex graph for a work-related task.

ChatGPT’s image recognition ability has already been tested by a company called Be My Eyes, which makes an app for people with impaired vision. Google has a product called Lens that can “see” an image and provide information about the content, so this feature isn’t a 100 percent new development.

The ChatGPT update is being rolled out to Plus and Enterprise customers now.

This image has an empty alt attribute; its file name is Firefly-question-marks-rising-from-flames-94723-1-300x300.jpg

What the Actual Hell?

Get ready for Snoop Dog, Paris Hilton, Tom Brady, and other celebs to become a face in your messaging app. Seriously. Among the slew of AI-related products that Meta showed off this week is an AI cast of characters featuring some familiar faces.

According to a press release, Meta “partnered with cultural icons and influencers to play and embody some of these AIs.” That’s right. Actual celebs were involved in creating these AI characters.

Not only that, “They’ll each have profiles on Instagram and Facebook, so you can explore what they’re all about,” according to Meta.

Where can you interact with Snoop Dog AI? You’ll be able to message him along with the others (28 in all) on WhatsApp, Messenger, and Instagram. Y’all do know these aren’t real, right? You won’t be actually taking to Snoop. Right? RIGHT? -— Michelle Johnson

A sampling of Meta’s AI characters. (Meta)

Fun and Useful Stuff to Try

No links this week. Here are some suggestions of AI-related things that you can try. -— Michelle Johnson

Put a Chatbot on Your Phone

Do you only use ChatGPT on your computer? Have you ONLY used ChatGPT? Well, it’s high time to try something new.

These four are available as mobile apps: ChatGPT, Bing, Bard, and Perplexity. Download them from your app store (iOS or Android) and experiment!

What about Google’s Bard? Sorry, no app yet. Use your phone’s browser. Or, follow Bard’s recommendation: “Personally, I recommend the Bing iOS app, as it gives you the best of both worlds. You can use it to search the web using the Bing search engine, and you can also use it to chat with ChatGPT and get answers to your questions in a comprehensive and informative way.”

So, when did Bard get personal? And it’s right. Bing features ChatGPT.

Tell ChatGPT and Bard: Check Yo’self!

How sure is that bot about what it told you? Find out by asking it to create a confidence score in a prompt. Example prompt, ChatGPT: Tell me the current population of Boston and provide a confidence score on the accuracy of the data. (Note: I tried this, and it gave me a 90% confidence score because “population figures can change over time and may not be exact.” It also provided a link to the source of the data.

On to Bard, which has a button (The big “G”) that shows up just below a response. If you’re giving Bard the side-eye, check up on it by clicking the G. Bard will then run a check on itself and color code anything that’s important. If something is highlighted green, it’s likely good to go. (I would STILL check it, tho.) If it’s brown, you should definitely do some further digging. It may not be correct.

Finally, Make Some Music

Get your creativity on. Hit this site, type in a prompt, and make some music here:

Aht Gallery

This week’s theme: Starry Night
(Michelle Johnson via Midjourney, DreamStudio, and LeonardoAI)

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.