Anthropic's Premium Chatbot

Claude Pro's enhanced features, including 5x more usage, priority access, and exclusive previews of new capabilities

Author

Wes Roth
September 08, 2023

Today:

Anthropic’s Claude AI chatbot gets a paid plan for heavy users

Anthropic, an AI company started by some former OpenAI folks, rolled out a paid version of their chatbot, Claude 2. For 20 bucks a month in the U.S., users get more bang for their buck - five times the usage of the free version, more messaging, front-of-the-line access during busy times, and a first look at new stuff. That price tag matches OpenAI's ChatGPT Plus, which is kinda like Claude 2's competition.

Folks like Claude for its snappy replies and smarts. With the pro version, you can send up to 100 messages every 8 hours, which is more generous than ChatGPT Plus. But there's a catch. If you're a chatterbox or send big files you'll use up your message limit faster.

Anthropic aims to make a super smart AI that can learn by itself, helping with tasks like answering emails and making art. Since starting in 2021, they've pulled in a whopping $1.45 billion in funds but think they'll need $5 billion in the next couple of years to make their dream AI. Most of that dough, especially from their Claude Pro sales, is gonna go towards the massive computer power they need. They're in the game with big names like OpenAI, and they're aiming high.

IBM rolls out new generative AI features and models

IBM's stepping up its game in the hot AI market with new features and models on its Watsonx platform. They've got these new "Granite series" models that are kinda like OpenAI's GPT-4. IBM says they'll spill the beans on how they trained these models later this year.

They're also rolling out tools to make these AI models work better for businesses. There's Tuning Studio, which lets you tweak models to do specific tasks, and a synthetic data generator for improving model training while keeping risks low (though what "low risk" means is still a bit foggy).

Soon, they'll add a chatbot-like tool to help folks interact with their data more easily. And they're also launching a toolkit for making sure AI models play nice with privacy and ethics rules.

Microsoft pledges legal protection for AI-generated copyright breaches

Microsoft's got your back if you get into hot water over copyright issues using their AI software like GitHub Copilot and Microsoft 365 Copilot. If someone sues you for copyright infringement because of something the AI created, Microsoft says they'll cover the legal costs.

This move eases worries for businesses thinking about using Microsoft's AI tools but scared of legal headaches. Adobe pulled a similar move earlier, all aiming to make folks more comfortable using AI that could accidentally copy someone else's work.

Microsoft's only got this offer for paying customers and says you've gotta use their built-in safety features for the deal to apply. Legal experts think Microsoft's risk here isn't too big, given how laws around AI and copyright are still a gray area.

Opera browser's 'Aria' AI chatbot can write custom bios and give tips on how to launch a streaming career

Opera's got a new toy for gamers: an AI chatbot named 'Aria'. Built into their GX browser, Aria dishes out game updates and even advice on how to start streaming. Powered by Opera's special tech and some OpenAI magic, Aria could learn more tricks in the future. You can ask stuff like "Is Starfield out yet?" or get a personalized character bio.

It's available worldwide, not just on PCs but also on Android and iOS devices. But, no worries if you're not into AI helpers; you've got to choose to turn Aria on. Wanna try it out? Flip on 'Early Bird' in settings, activate a couple of 'Aria' options, and then sign in. Also, a neat update for the Opera GX browser: a quick keyboard command to chat with Aria and a cool AI Prompts feature.

Morgan Stanley to launch AI chatbot to woo wealthy

Morgan Stanley's getting into the AI game to jazz up how it deals with rich clients. They've been testing a chatbot—made with the help of OpenAI, the folks behind me—and it's about to roll out. This isn't replacing your human financial adviser; it's more like a sidekick that can quickly pull up research, fill out forms, and even jot down meeting notes. Morgan Stanley thinks this could be as big of a game-changer as the internet was back in the day.

This AI push isn't just a cool gadget. It's part of a bigger plan to grow Morgan Stanley's wealth management business, which has been doing pretty darn well lately. And they're not the only ones. Big banks like JPMorgan Chase and Bank of America are also messing around with AI to make their services better.

Mindtrip wants to become your AI travel agent

Mindtrip wants to be the travel agent you never knew you missed. Back in the day, folks used real-life agents to plan their trips, but the internet pretty much made them history. Now a group of tech-savvy entrepreneurs sees a hole in the game—making trip planning easy-peasy but without the human middleman. So they've cooked up Mindtrip, an AI that'll plan and book your trip like a pro. They just raised $7 million to get this ball rolling.

Andy Moss, the guy steering the ship, has a track record of building and selling companies. He's joined by some other smart cookies, including tech veterans from Google. Mindtrip uses everyday language to help you plan your trip, showing you pictures and even helping you book stuff.

AI RESEARCH LAB IMBUE NABS $200 MILLION FOR SPECULATIVE BET TO BUILD AI ‘AGENTS’

Two entrepreneurs, Kanjun Qiu and Josh Albrecht, who used to dream about launching their own AI lab, finally did it. Their startup, Imbue, just snagged a fat $200 million investment, led by Jed McCaleb, the crypto billionaire. Imbue's now valued at over a billion bucks, and it's one of the rare woman-led AI unicorns out there. They're not planning to build the usual chatbots; they're aiming for AI "agents" that can do more complex tasks all by themselves, like a virtual research assistant on steroids.

Imbue is working with a lot of computing power, similar to what OpenAI used for creating GPT-3. But don't hold your breath; they don't have a demo yet and say it'll be a while before we see a finished product. The big investment is risky 'cause they don't have any revenue, but they're hoping to break new ground in AI. The goal? To make AI so user-friendly it's like having a truly personal computer that does the heavy lifting for you.

Even though they're kinda outsiders in the AI game, they're not sweating the skeptics. They're playing the long game, taking their sweet time to make sure what they put out there is not just good, but great. So while we won't see anything from Imbue in the next few months, they're one startup to keep an eye on.

AI reading coach startup Ello raises $15M to bolster child literacy

Ello, a startup focused on wiping out childhood illiteracy, just bagged $15 million in funding. Using AI, the service listens to kids read and helps 'em out with tricky words and pronunciation. For $24.99 a month, your kiddo gets five books selected just for them, and the app is fine-tuned to their reading level and interests.

Founded by experts from Google and Stanford, Ello is different from other edu-apps because it's more about making reading fun and less about stressful tests. The company already has 10,000 families on board and plans to roll into schools soon. So, not only is Ello helping parents, but they're also aiming to give teachers a helping hand.

Facebook Trains Its AI on Your Data. Opting Out May Be Futile

Facebook's parent, Meta, is beefing up its AI with your personal info. Zuckerberg's pushing hard into AI territory with features for Facebook, Instagram, and all the gang. And while they've slapped a new form online for you to ask about any third-party data they've got on you, don't hold your breath. First off, this ain't about the mountain of info they've already hoovered up from their platforms—it's just the extra stuff from other corners of the web or bought from data peddlers.

Want to try the form? You'll need your country, name, and email. Heads up, it's a pain on mobile. And even after you hit submit, you might be left hanging. Folks across the pond in the UK might have a better shot because of stricter privacy laws, but for us Yanks? Who knows. By the way, this move by Meta? Comes on the heels of them getting slapped with a fat $1.3 billion fine in Europe for playing fast and loose with data. So, keep your eyes peeled and your data closer.

Low-code platform Retool makes it easier to bring AI smarts to business apps

Retool, a startup focused on making business apps, is stepping up its game by adding AI capabilities. The company's boss, David Hsu, says businesses want to add AI to their apps but struggle with keeping data updated. So, Retool's launching a new service called Retool Vectors, which is like a storage locker for the data AI needs. It keeps that data fresh by syncing with other systems like Salesforce.

This helps businesses use AI more effectively without spending a ton of money or time. Retool tested this out with their own chatbot and saw their customer help success rate jump from 20% to 60%. The company's also rolling out some new AI features, like text summarization, made in collaboration with OpenAI. The idea is to make it easier and quicker for businesses to get the benefits of AI.

How AI may be a powerful tool in treating male infertility

Dr. Steven Vasilescu and his team have whipped up an AI system called SpermSearch to help find healthy sperm in samples from guys who can't naturally father a child. This thing is like a bloodhound for sperm; it's 1,000 times faster than a human expert looking through a microscope. So, instead of spending hours searching for a 'needle in a haystack,' the software spots the good stuff in seconds.

But hold your horses, this isn't replacing human experts. It's more like their sidekick, helping to speed up the process. That's a big deal because when you're trying to fertilize an egg, the clock's ticking.

While the tech is promising, it's still in the early stages. It could be a few years before we see it in action, and it's mostly aimed at a small group of guys with a specific fertility issue. Still, if it works out, it could be a game-changer for couples struggling to have a kid.

TIME Reveals Inaugural TIME100 AI List of the World's Most Influential People in Artificial Intelligence

TIME just dropped their first-ever list of the top 100 movers and shakers in the AI world. Think of it as a who's who of big brains making waves in artificial intelligence. Names like Sam Altman of OpenAI and Demis Hassabis of Google DeepMind are on there, plus artists, scientists, CEOs—you name it.

They even got some young blood, like 18-year-old Sneha Revanur, and some seasoned pros, like 76-year-old Geoffrey Hinton. And it's not just a boys club; 41 women and nonbinary folks made the cut too. TIME picked these people because they're setting the stage for where AI is headed. Oh, and keep an eye out—TIME's got a bunch of events coming up that'll focus on women in AI.

What'd you think of today's edition?

Login or Subscribe to participate in polls.

What are MOST interested in learning about AI?

What stories or resources will be most interesting for you to hear about?

Login or Subscribe to participate in polls.

Reply

or to participate.