OpenAI gives GPT-4 a major upgrade

Read time: under 4 minutes

Welcome back, Superhuman

The AI-focused Cloud Next conference from Alphabet brought a list of major announcements. But OpenAI had one or two major announcements of its own.

TODAY’S MENU

• All the big announcements at the Cloud Next event

• How to use Gemini 1.5 Pro

• Prompt of the day

• A new way to discover new music with AI

• 5 new AI tools to boost your productivity

• AI Generated Images: A 1960s vision of the future

NEWS

Today in AI & Tech

• Show, Don’t Tell: OpenAI is rolling out a new version of GPT-4 Turbo with Vision — the ability to understand and analyze images — to its API users.

• Upgraded: OpenAI also said they’re rolling out an upgraded version for GPT-4 to ChatGPT.

• Thinking Ahead: In a recent interview, Elon Musk speculated AI will be smarter than any human on earth by the end of 2025 — a jump from his last prediction of 2029.

• PPM: Poe, the AI chatbot platform owned by Quora, is introducing a new revenue model for its creators, allowing them to set a per-message price for their bots.

• Future Scan: German healthcare company Bayer and Alphabet’s GCP are creating an AI-powered platform that’ll help radiologists diagnose patients by flagging anomalies in scans. We have more on Alphabet’s Cloud Next event below.

NEXT IN AI

A guide to the big announcements at the AI-focused Cloud Next event

AI was the star of the show at the search giant’s Cloud Next event. The biggest piece of news might be that the tech giant is going to release a custom AI chip later this year that’s built on Arm technology. That’s the same chip architecture that Microsoft and Amazon Web Services have been using for years. But Alphabet says its version’s performance is about 30% better.

That was just the tip of the iceberg. The company made a number of major announcements that could shake up the AI leaderboards for years to come, including…

• Vids, a tool that lets business leaders create short video clips with AI — kind of like PowerPoint for videos

• Three open-source tools that firms can use to build new AI-powered projects

• An update to Gemini 1.5 Pro that lets the platform understand and analyze audio (It’ll now be able to summarize earnings calls or compare different radio broadcasts, for example)

• Code Assistant, a competitor to Github’s popular Copilot platform that offers coding suggestions and autocomplete tools for developers

• Updates for its text-to-image generator, Imagen 2, including a feature that lets you turn text into an animated GIF

• A $10 per month Workspace subscription that gives users access to AI-powered meetings, messaging, and translations

PRESENTED BY SANA

Your AI assistant for work

Sana AI is a knowledge assistant that helps you work faster and smarter. You can use it for everything from analyzing documents and drafting reports to finding information and automating repetitive tasks.

Integrated with your apps, capable of understanding meetings and completing actions in other tools, Sana AI is the most powerful assistant on the market.

Try it here

AI AT WORK

Gemini 1.5 Pro is now available in over 180 regions — how to access and use it

After a limited release a few weeks back, Gemini 1.5 Pro is now available in 180+ countries. It comes with a larger context window and expanded token limit, which means you can upload very large files like entire books and prompt them.

It also comes with image, video, and audio analysis, as well as memory recall that’s almost perfect — making it the best AI model in theory for extracting accurate details from large files.

Here’s how to use it:

• Go to the AI Studio website

• Select Gemini 1.5 Pro from the Model menu on the right

• Attach a file from the Insert menu in the top of the window

• Type a prompt in the chat and you’re good to go

While you can do all the regular stuff here that you can do with other chatbots like ChatGPT, Gemini 1.5 Pro really shines when you upload large files and try to get answers or summaries related to the file.

Note: AI Studio can be accessed in most major countries but it’s yet to be rolled out in a small number of locations. If you see a list of available countries when you visit the AI Studio website, it means AI Studio is not available in your region.

PROMPT OF THE DAY

Create high-converting website copy

Prompt: I am creating a landing page about [insert detail]. The goal of the landing page is to [insert a goal, like increasing installs]. Below is a description of the product and the target audience. Generate high-converting copy for the website: [Share details]

PRESENTED BY WEBAI

Securely and efficiently test & build AI solutions

With lower data requirements, local training, and a no-code environment, we’ve de-risked the creation and deployment of AI solutions. Now you can spend more time iterating on ideas and less time worrying if they will work. Check it out here

AI & MUSIC

Spotify’s next frontier: AI-powered playlists

These days, there’s a playlist for everything. Well, not everything. Spotify already has options based on your mood or the time of day — like “sunny day pop music” or “laid-back, stress-free songs.” But what if you want to get more specific than that? Say, a playlist based on the movie you just watched — or a painting you’re working on.

ChatGPT for music: Spotify says it’s rolling out a new AI feature that will let premium users do just that. Premium users in the UK and Australia will soon be able to type in a text prompt based on “places, animals, activities, movie characters, colors, even emojis” — and get a custom, 30-song playlist.

What’s next? If the beta goes well, users in the US and the rest of the world can probably expect to get their hands on the feature in the coming months. Spotify’s big competitor, Apple Music, might have some AI features in the works, too. We’ll likely hear more about its AI innovations at its Worldwide Developers Conference in June.

PRODUCTIVITY

5 AI Tools to Supercharge Your Productivity

Sourcegraph: Rapidly search, write, and understand code, using insights from your entire codebase.

Brave Leo AI: Privately ask questions, summarize pages, and create new content with a smart AI assistant built into our browser.

MindStudio*: Unlock the power of AI without any coding. Join a free webinar on 04/16 at 1 PM PT and learn how to build custom AI apps for your team. Reserve your spot now!

Odaptos: Explore users’ feelings and emotions, enhance their user experience, and fast-track your user test process with AI.

AssemblyAI: Turn calls, virtual meetings, and podcasts into insights, transcripts, and summaries.

^{* indicates a promoted tool, if any}

AI GENERATED IMAGES

A 1960s vision of the future

Prompt: A Retrofuturism-inspired photo capturing a 1960s vision of the future, where cars run on green olive fuel, and architecture incorporates olive shapes in its design, reflecting the optimistic futurism and artistic signature of Dan McPharlin.

Acquire new customers and drive revenue by partnering with us

Superhuman is the world’s biggest AI newsletter for businesses and professionals with 600,000+ readers working at the world’s leading startups and enterprises. Companies like Amazon, Calendly, and Notion feature their products in Superhuman. Main ads are typically sold out 4 weeks in advance. You can book future ad spots here.  

🧞 Your wish is my command

Thanks for reading.

Until next time!

Zain

p.s. If you liked this newsletter, share it with your friends and colleagues here. For questions and feedback, email us at [email protected]. We always listen and respond.