Apple's New Venture: AI-Powered Health Coaching to Change Your Life

Today:

Apple Is Reportedly Developing An AI-Powered Health Coaching Service

Well, well, well, Apple's at it again, folks. This time, they're reportedly putting together a health coaching service, called Quartz for now, that's got more brains than a MENSA convention. You know, the AI stuff. The whole shebang's designed to nudge you off the couch, eat a salad once in a while, and maybe even get a decent night's sleep.

And that's not all. The company's tech wizards are supposedly cooking up tools to track your moods and manage vision issues, like myopia - or nearsightedness if you're not down with the fancy lingo. They're also planning to launch an iPad version of the iPhone Health app, cause who doesn't want to see their heart's rhythm on a big screen, right?

They're hoping to roll out this health coaching service next year, but hey, the tech world is as unpredictable as my Aunt Mabel's meatloaf - they could delay it, or even shelve it completely.

Apparently, Apple's also been dreaming of adding meditation features to its upcoming mixed-reality headset - perfect for all you Zen seekers out there. Also on the drawing board are basic blood pressure monitoring for the Apple Watch, and a new method to measure your blood sugar without the usual jab to the finger. Though the latter's still in the works, Apple hopes to stuff this tech into its watch eventually.

Now, don't forget, this ain't Apple's first rodeo in health tech. They've been at it since 2014, adding all sorts of health-related gimmicks to their smartwatch over the years. So, keep an eye on this space - Apple's health tech is shaping up to be as exciting as a double espresso on a Monday morning!

READ THE ARTICLE ON TECHCRUNCH.

NVIDIA Open-Source Software helps Developers Add Guardrails To AI Chatbots

In the wild west of AI chatbots, NVIDIA's just given developers a new sheriff in town. This open-source software, NeMo Guardrails, keeps AI applications from going off the rails, ensuring they're accurate, appropriate, on topic, and secure. No more AI chats turning into a wild goose chase!

All kinds of industries are hitching their wagons to large language models (LLMs), the horsepower behind AI apps. They're doing everything from answering customer queries to writing software and speeding up drug design.

NeMo Guardrails is like a well-trained shepherd dog, keeping all these AI-powered applications in line. It works with every LLM, including OpenAI’s ChatGPT, and helps developers keep their apps safe and within their area of expertise.

It sets up three types of boundaries:

• Topical guardrails: Keeping your AI from chattering about irrelevant stuff, like a customer service bot suddenly giving weather updates

• Safety guardrails: Making sure the app's responses are accurate, appropriate, and backed by credible sources.

• Security guardrails: Only letting apps connect to external applications that are known to be safe.

The good news? You don't need to be a machine learning whiz or data scientist to use it. A few lines of code and voila! New rules set up.

NeMo Guardrails can work with all the tools that developers use, like LangChain. Harrison Chase, the creator of LangChain, says you can add NeMo Guardrails to LangChain workflows to quickly set safe boundaries for your AI apps. It also works with a wide range of LLM-enabled apps like Zapier, an automation platform used by 2 million businesses.

NVIDIA has integrated NeMo Guardrails into the NVIDIA NeMo framework, which includes everything you need to train language models. This framework is available as open-source code on GitHub, as a supported package from NVIDIA, or as a service. South Korea's leading mobile operator has already used it to build an intelligent assistant that's had 8 million convos with its customers. A Swedish research team used NeMo to automate text functions for hospitals, government, and business offices.

Building good guardrails for AI is a tough cookie, needing continuous research as AI evolves. NVIDIA's thrown its hat in the ring with NeMo Guardrails, making it open source to contribute to the developer community’s work on AI safety. The goal? To help companies keep their smart services safe, private, and secure, so these engines of innovation stay on track. That's the ticket! For more details, check out their technical blog.

READ THE ARTICLE ON NVIDIA.

Salesforce Announces Einstein GPT For Field Service

Well, butter my biscuit! Salesforce is adding some fancy gizmo called Einstein GPT to its Field Service app. Now, hold your horses, it ain't a new type of GPS, it's actually artificial intelligence (AI), designed to help folks in field service jobs do their tasks better. We're talking home nurses, technicians, contractors, and the likes.

The new AI addition is like a handyman's helper, making things like note-taking, communication, and service reporting less of a chore. It's a bit like having a high-tech personal assistant that doesn't need coffee breaks.

Working with OpenAI, Cohere, and Anthropic, Salesforce aims to make the Field Service app more efficient, manage tasks and assets, schedule work, and improve customer service. The addition is particularly helpful for part-time contractors. It's like having a smart calendar that knows when these folks are free and can work.

There's this cool thing they're calling "service swarming". Imagine a bunch of folks in your company jumping on Slack to tackle an issue together. It's like a virtual huddle, only smarter.

This Einstein GPT also helps with on-the-job training and making sure everyone knows what's what before they start a job. Say, a technician knows what the previous guy did at the same site, so there's no reinventing the wheel. It's also like having a savvy salesperson in your pocket, suggesting other products a customer might like.

The AI also provides step-by-step guides, tailored instructions, and even takes into account public data and data from your own system. It's like a smart encyclopedia crossed with a GPS that you can ask anything.

But don't get ahead of yourself. This whole AI thing in field service is just getting started. A lot of smaller companies are still playing catch up with all this digital stuff. Salesforce sees a future where AI enhances field service jobs rather than replacing them. It's like having a super-smart apprentice working alongside humans, making everything more efficient.

In the long run, Salesforce is looking to combine their mobile apps' functions to provide a smoother user experience. There are also enhancements in the pipeline for their Data Cloud and Flex Worker Management, aiming to prevent machinery failure and optimize field worker dispatch.

So, it looks like Salesforce is really putting the pedal to the metal with this AI thing. Could be a real game-changer in the field service world. Stay tuned for more updates. And remember, don't let the tech jargon scare you off. It's all just fancy ways of saying "we're making things easier for you".

READ THE ARTICLE ON TECHREPUBLIC.

Yelp Enhances Consumer Experience With New Discovery, Review And Services Features

Well, Yelp's pulling out all the stops, folks! They're jazzing up the joint with some nifty new features aimed at making our lives a smidgen easier.

First up, they're introducing "Yelp Guaranteed." It's like an insurance policy for hiring home service pros. If your plumber turns your bathroom into a water park, Yelp's got your back, promising up to $2,500 if things go haywire. It's currently only available in a few major cities, but they're planning on rolling it out across the country this summer. Nice!

Yelp's also flexing its techie muscles, using some fancy-pants AI to make searching for businesses more intuitive. No more sifting through unrelated search results when you're hunting for a specific need. They're even including helpful review highlights right under each business listing.

On the topic of food, they've got a new "Surprise Me" feature that suggests highly-rated restaurants when you're being indecisive. And for the times when you're feeling a bit of writer's block after a meal, they've got these new interactive review topics like "food," "service," and "ambiance" to help jog your memory.

Adding a bit of fun to the mix, they've introduced new reactions to reviews. You can now show your appreciation with "Helpful," "Thanks," "Love this," and "Oh no" reactions.

They're also letting businesses in on the action. Businesses can now upload videos to their posts, so you can get a sneak peek of a new dish or upcoming events.

Navigating their pages should be a cinch now, with a new navigation bar making it easier to find what you need. And let's not forget about the photo updates, making it easier to find the most mouth-watering menu photos.

Finally, they're ditching the need for password memory with a new login experience. You'll get an email link that'll log you in with one click.

So, there you have it. Yelp's making some major moves, all with the goal of making our lives just a tad bit easier. Stay tuned for more updates, folks!

READ THE ARTICLE ON YELP.

Arize Launches Phoenix, An Open-Source Library To Monitor LLM Hallucinations

You know how it is when your AI chatbot starts seeing pink elephants? That's called hallucination, and it ain't good, especially when we're talking about legal and medical data. Enter Arize AI, the folks from sunny California. They've just whipped up Phoenix, a tool that keeps an eye on these big, complex language models so they don't start making things up.

These language models are the hot new thing. They're behind all sorts of apps, like chatbots offering legal advice or summarizing medical appointments. But as slick as they are, they can still goof up and spit out false or misleading info. Phoenix is built to spot these missteps and flag 'em up for review.

Here's how it works. You feed Phoenix a chatbot convo, and it studies the data to see where the bot did well and where it tripped up. Then it'll show you all that in a nice, neat visualization. If there's a problem – like your bot flunking Spanish 101 – you can dive in and fix things up.

The Arize folks reckon Phoenix is the first tool of its kind built to evaluate and manage these language models, but don't forget that the tech world moves quicker than a jackrabbit on a hot griddle. There's always someone new ready to join the party.

Still, the early reviews for Phoenix are as sunny as a California beach. Christopher Brown, a bigwig from Decision Patterns, said it's a big step forward in understanding and improving these models. And let's face it, anything that helps stop your AI from seeing pink elephants has gotta be good, right?

READ THE ARTICLE ON VENTUREBEAT.