What really happened at OpenAI

ALSO: These ChatGPT alternatives are getting really good

Read time: 2.5 minutes

Welcome back, Superhuman

Another day, another bombshell revelation coming out of OpenAI. This time Reuters and the Information claim that OpenAI might have made an AI breakthrough so powerful that it spooked the board and led to CEO Sam Altman’s firing.

While the chaos at OpenAI continues, ChatGPT alternatives are making significant progress, with powerful new upgrades and tens of millions of new users. We round up what’s new, and how you can use it.

TODAY’S MENU

  • New reports suggest OpenAI made a shocking AI breakthrough

  • ChatGPT alternatives are getting really good

  • Research: AI models can hide their reasoning from humans

  • Tutorial: Create catalog photography in ChatGPT

  • AI Generated Image: Lord of the Gangster Rings

TODAY IN AI & TECH

Source: Google

  • Video Watcher: Google’s Bard chatbot can now process YouTube videos and answer questions about it. Useful for summaries, analysis and more.

  • Bad for Business: Almost $1 Billion flows out of crypto exchange Binance after its CEO pleads guilty to money laundering charges.

  • Back on Track: OpenAI’s share sale at $86 Billion valuation is now reportedly set to continue after the return of CEO Sam Altman.

  • Movie Maker: Stability AI introduces Stable Diffusion Video, a new AI model that generates videos.

  • Rocketship: Nvidia announces $18 Billion in revenue for the previous quarter, tripling what they earned for the same quarter last year.

THE CHAOS AT OPENAI

Reports claim OpenAI staff warned board of AI breakthrough ahead of CEO Sam Altman’s firing

Image source: OpenAI

Just when you think things can’t get any crazier, new reports from Reuters and The Information claim that researchers at OpenAI warned the company’s board of an AI discovery so powerful that it could “threaten humanity.“

The reports claim that the letter that revealed this information was one of the key factors that led to CEO Sam Altman’s firing. Employees who wrote the letter reportedly feared that the company would attempt to commercialize this technological breakthrough before fully understanding the consequences.

According to Reuters, the project at the center of the controversy is called Q* (pronounced Q-Star), a discovery that could lead to artificial general intelligence (AGI), an AI system that could surpass human capabilities.

These reports could be in line with a recent statement from Sam Altman where he said: “Four times now in the history of OpenAI, the most recent time was just in the last couple weeks, I've gotten to be in the room, when we sort of push the veil of ignorance back and the frontier of discovery forward, and getting to do that is the professional honor of a lifetime.“

IN PARTNERSHIP WITH GUIDDE

Explain and document the most complex tasks in seconds with AI

We all have that one colleague who keeps asking the same thing over and over again, no matter how many times we explain it to them.

It’s time you let AI do the explaining instead of you. Guidde is an AI-powered tool that helps you explain the most complex tasks in seconds with AI generated documentation:

  • Turn boring documentation into stunning visual guides

  • Save valuable time by creating video documentation 11x faster

  • Use it to document workflows for your teammates, share insights across your company, train and onboard new hires, and much more

Simply click capture on our browser extension and the app will automatically generate step-by-step video guides complete with visuals, voiceover and call to actions.

Guidde is used and trusted by 20,000+ users. And we’re rated 5/5 stars on the Google Chrome store. Try it here at no cost

AI AT WORK

ChatGPT alternatives are getting really good. Here’s what’s new and how you can use it:

Source: Anthropic

There’s a long list of ChatGPT rivals competing to take the throne of chatbot king — and some are finally starting to build a case for being a genuine alternative.

Here’s a brief overview of what’s changed recently, and why you might want to try these alternatives:

Anthropic releases Claude 2.1

Perhaps the best alternative to ChatGPT, Anthropic just released an upgraded version of their Claude chatbot. The new Claude 2.1 can now process prompts of up to 150,000 words or 500 pages. This means that you can upload entire books, reports, code bases and large datasets.

The new model also comes with a 2x decrease in incorrect answers. These upgrades make Claude a compelling alternative for users who want to work with longer documents and larger file sizes.

Grok is coming to premium X users next week

While the accuracy of Elon Musk’s Grok doesn’t match that of ChatGPT and GPT-4, the chatbot has one unique advantage over ChatGPT: It has access to real-time information from X, unlike ChatGPT which only has access to information up until April 2023.

This feature could provide a handy use case for readers who need access to the latest information to generate the output they need. The feature will be available to all premium users next week, according to Musk.

Baidu’s Earnie 4.0 chatbot reaches 70 million users

Earnie 4.0 is Chinese tech giant Baidu’s answer to ChatGPT. The company claims that its chatbot performs just as well as ChatGPT. And according to a recent announcement from the company, Earnie 4.0 has now reached 70 million users in just 3 months — astonishing growth for any product. While ChatGPT is still the king of the hill, it could see some stiff competition coming from Chinese alternatives in the coming months.

BUZZWORD BUSTER

Effective Accelerationism

Accelerationism is an ideology that advocates for social transformation through drastic intensification of capitalist growth and technological change. Effective Accelerationism is an extension of this ideology that calls for progress at all costs.

This school of thought has recently gained some traction in Silicon Valley, especially in the field of AI, where some believe that artificial general intelligence could unlock transformational progress.

5 AI Tools to Supercharge Your Productivity

Bardeen: Delegate your tedious work with a few lines of text. Explore integrations with your favorite apps and hundreds of pre-built playbooks.

Dubecos: Dub your videos effortlessly using AI.

Humanlinker: Helps revenue generation teams skyrocket past quotas with personalized prospecting and efficient meeting preparation.

Deepgram: Build speech AI into your apps with this text-to-speech API that transcribes and understands audio. Used by NASA, Spotify and more.

PicWonderful: Generate images from text. Remove backgrounds, un-blur photos, or colorize them.

RESEARCH

AI language models can use steganography to hide their reasoning

Redwood Research, a research lab for AI alignment, has discovered that LLMs are capable of concealing their thought processes using a technique called steganography.

The finding implies that AI models can discreetly store the steps they take to think about an issue inside the words or the way sentences are phrased, which makes it difficult for humans to grasp how the AI is thinking or to test its performance. This finding has significant implications for regulating AI systems and ensuring their transparency and trustworthiness.

TUTORIAL

How create catalogue photos in ChatGPT

Want to create highly professional photos for your product’s sales catalog? Use the prompt below:

Prompt: Photograph [insert object] for a sales catalog, ensuring the [insert object] is the sole focus with no additional visual elements. Center the [insert object], fully visible, in high detail to emphasize its design and quality. Use a plain grey background for a clean and professional look. Focus on natural, clear lighting to highlight the [insert object] features.

We’ve tried the prompt on both ChatGPT and Midjourney and generated equally good results.

source: .teddm on Midjourney

AI GENERATED IMAGES

Lord of the Gangster Rings

Source: @lotrspecticle on Midjourney

📈 Feature your product in the world’s biggest AI newsletter

Superhuman is the world’s biggest and fastest-growing AI newsletter with 450,000+ readers working at the world’s leading startups and enterprises. Companies like Masterworks, Brave, SEMRush, and 1Password feature their products in Superhuman. Main ads are sold out for the next 4 weeks. You can book future ad spots here.  

🧞 Your wish is my command

What did you think of today's email?

Your feedback helps me create better emails for you!

Login or Subscribe to participate in polls.

Reviews of the day

Thanks for reading.

Until next time!

p.s. if you want to sign up for this newsletter or share it with a friend or colleague, you can find us here