Gen-3 Alpha is Available, $27,1B for AI Startups, and Figma Disables AI Features
PLUS HOT AI Tools & Tutorials
Hello and welcome to our weekly digest!
Don't you get the feeling that AI encourages procrastination? Whenever I think about mastering video editing, the voice in my head says, "Nah, a couple more months, and I can do the same thing with a single prompt." It makes me think we're heading towards a WALL-E scenario, not The Terminator.
In fact, today's featured material is about just that. We also have plenty of AI news on other interesting topics. So, let's think positive thoughts and get started!
And if you want to read about people who are not procrastinating, but actually doing something useful for humanity, this is the newsletter for you:
This Creators’ AI Edition:
Featured Materials 🎟️
News of the week 🌍
Useful tools ⚒️
Weekly Guides 📕
AI Meme of the Week 🤡
AI Tweet of the Week 🐦
(Bonus) Materials 🎁
Featured Material 🎟️
You Can Try Runway Gen-3 Alpha Now
RunwayLM (one of the first AI developers for video generation, by the way) has unveiled the public version of a new Gen-3 Alpha model. In case you missed the announcement, I will remind you that this is the first of an upcoming series of models trained by Runway on a new infrastructure built for large-scale multimodal training. The startup said it significantly improves fidelity, consistency, and motion over Gen-2.
A few weeks after the announcement, everyone can test the Gen-3 Alpha. With a small disclaimer: the new model is available only to paid subscribers starting at $15/mo (or $12/mo if you pay for an annual subscription).
Given the huge demand for models like OpenAI's Sora and Luma's Dream Machine, the decision to make Runaway's new AI paid seems obvious. In addition to wanting to monetize their product, the startup needs to keep their data centers stable. With a free platform, this would have been difficult.
Keep your mailbox updated with key knowledge & news from the AI industry
Gen-3 Alpha supports all the major modes from Runaway. These include text-to-video, image-to-video, and text-to-image, as well as existing tools such as Motion Brush, Advanced Camera Controls, and Director Mode. In addition, the model has been released with a new set of security measures, including an improved in-house visual moderation system and C2PA provenance standards.
Gen-3 Alpha Features
You can find dozens of AI-generated videos on the Runaway website.
The first tests show that the model handles complex and realistic scenes well. People look real, and various details retain physics from the real world. This is clearly visible in the close-up videos. Take a look at the example below.
Prompt: A close up portrait of a woman lit by the side, the camera pulls back
With other styles, Gen-3 Alpha also looks promising. You can ask the AI to generate an anime-style character or request a very abstract video. Runaway noted that companies interested in tweaking and creating customized models can contact the startup.
This fact suggests that we'll see new and more sophisticated iterations of Gen-3 soon.
Prompt: A Japanese animated film of a young woman standing on a ship and looking back at camera.
Of course, Twitter is already flooded with posts about what we can come up with with the new AI. And some of the stuff, in the internet tradition, looks very creepy. To see what I'm talking about, just type "Runway" or "Gen-3 Alpha" into a search bar.
I wonder how many more times this year I'm going to say this is the year of integrating AI into video content production.... Anyway, OpenAI, Luma AI and Runway, tell me, when can we generate a satisfying ending for Game of Thrones?
Yeah, I still haven't let it go.
News Of The Week 🌍
PitchBook: AI Startups Raised $27,1B for Q2 2024
According to PitchBook, investors poured $27.1 billion into AI startups between April and June, accounting for nearly half of all U.S. startup funding. In total, startups in the U.S. raised $56 billion, up 57% from a year earlier and the largest three-month increase in two years. The total number of deals has also increased. Some analysts suggest that AI has restored investors' interest, and the decline of the VC industry is ending.
I can't say the PitchBook report is surprising. We're used to AI startups breaking records for the number and size of funding rounds. However, sometimes it's good to pay attention to news like this to understand where we stand. By the way, if you consider yourself an AI skeptic, check out this post. You'll like it:
Figma Disables Its AI Design Features
Last week, Figma introduced a massive update with generative AI features. The main part of the update was the "Make Design" tool, which creates mockups for new apps. As it turns out, these mockups aren't always unique. NotBoring Software founder Andy Allen noticed that when he was trying to create a template for a weather app, Make Design simply copied Apple's Weather.
Allen, in a Twitter post, accused Figma's creators of practicing AI on existing platforms. In response, CEO Dylan Field denied the allegations. But the company has still pulled the plug on Make Design, and it remains to be seen when the feature will be restored.
You may find it strange, but this news speaks much more eloquently about the hype around AI than the previous one. Companies are so afraid of missing the train that they make increasingly stupid mistakes. At the same time Figma developers are a long way from Google's failures.
Anthropic Will Fund The Next Generation of AI Benchmarks
Anthropic has announced the launch of a program to fund the development of new types of benchmarks capable of evaluating the performance and impact of AI models, including generative models such as its own Claude. Here are the highest priority focus areas for Anthropic:
AI Safety Level assessments
Advanced capability and safety metrics
Infrastructure, tools, and methods for developing evaluations
Our investment in these evaluations is intended to elevate the entire field of AI safety, providing valuable tools that benefit the whole ecosystem. Developing high-quality, safety-relevant evaluations remains challenging, and the demand is outpacing the supply.
I still have little faith in the usefulness of AI benchmarks for average users. There are dozens (if not hundreds) of different tests, but all are a poor indication of how the average person handles AI. However, benchmarks for B2B look like a really useful and important initiative. Let's wish Anthropic the best of luck.
Sharing is caring! Refer someone who recently started a learning journey in AI. Make them more productive and earn rewards!
Apple Poised to Get OpenAI Board Observer Role as Part of AI Pact
New details have emerged about Apple's partnership with OpenAI. Bloomberg reports, citing insiders, that Apple will get an observer role on OpenAI's board of directors as part of the agreement announced last month. They say the candidate responsible for the position has already been selected. Phil Schiller, Apple's App Store chief and former chief marketing officer, will take the chair.
Not only is ChatGPT making its way into iPhones, iPads and Macs, but Apple is also making its way into OpenAI. How this news will affect OpenAI remains to be seen, but it's interesting that Apple will now support Sam Altman's startup at the Microsoft level, which has long been a major backer and board member.
News Outlets Accuses Perplexity of Plagiarism and Unethical Web Scraping
Perplexity seems to be in trouble. Forbes accused the startup of plagiarizing one of its news articles in the beta version of Perplexity Pages. Wired has also blamed Perplexity for illegally copying its site and other sites. In addition, Amazon is conducting a similar investigation into the startup. Perplexity denies the accusations, says it honors publishers' requests not to copy content, and operates under copyright fair use laws.
TechCrunch pointed out that the uncertainty around copyright laws and AI-powered web scanners complicates the situation.
You can read a detailed breakdown here.
Useful Tools ⚒️
Hirebase - AI Job Search Engine
Galah - AI Video Ad Transcription & Analysis
Vitamin AI - Companion for Task Management
Motiff – AI-powered professional UI design tool
BrainyAI - Group Chat with Top-Tier AIs at Once
I'm not a fan of centralization and I'm sure the more diverse chatbots and models we have, the better. Nevertheless, constantly switching between tools is tiring and just plain annoying. BrainyAI solves this problem. This tool works as an extension for Chrome and gives you access to popular AI models like GPT-3.5, GPT-4, GPT-4o, Claude, Llama, Gemini, and etc. Everything you need in one chat, sounds great.
Share this post with friends, especially those interested in AI stories!
Weekly Guides 📕
Create 10 Engaging AI Shorts in 30 Seconds: Ultimate Guide to ClipGen AI
Master Midjourney - Website Full Guide, Every New Feature Explained
How I run a $25,000/month faceless YouTube channel with AI
Invideo AI - Full Tutorial 2024: Best AI Video Generator
AI Meme Of The Week 🤡
AI Tweet Of The Week
I can't wait to see these athletes in the Olympics!
(Bonus) Materials 🏆
Magic and Wonder in the Age of AI | David Kwong | TED
That Essay Got a B+. An AI Bot Graded It.
The A.I. Boom Has an Unlikely Early Winner: Wonky Consultants
Aravind Srinivas: Perplexity CEO on Future of AI, Search & the Internet
AI in Retail: How It’s Impacting the Industry & What You Need to Know
How was this week in AI? Share your content and ideas in the comments to this post so we can discuss or include them in the next edition!