Pika 1.5 Explodes, Luma AI gets 10x faster, Kling Talks, & Minimax Sees!
The Generative AI video space is rapidly evolving, with new features rolling out every week. We've been working hard to keep marketers up-to-date by summarising the latest tools and options available.
This week, we’re thrilled to showcase a series of new videos we’ve contributed to, adding musical elements to the legendary Dave Clark's remarkable projects. Dave, who joined us on our live podcast last month, has just released Battalion - a groundbreaking 5-minute AI short film that tells the story of a soldier from the only African American division to land on D-Day. Additionally, we’re excited to share Nike Equestrian 2045, a futuristic spec ad featuring our remix of Depeche Mode, alongside our remix of London Grammar in Battalion.
Top 5 AI things for marketers to know this week...
Pika Labs 1.5 update with built-in visual effects
Luma Labs introduces hyperfast video generations (10x faster)
MiniMax AI is rolling out image-to-video this week
Kling 1.5 introduces a new lipsync feature
California’s controversial AI safety bill vetoed
We went deep into all these updates during this week's ‘AI Live’ session, but for those who couldn’t make it you can watch it here.
Pika Labs 1.5 update with built-in visual effects
In a galaxy far, far away, Pika Labs 1.0 was a pioneer in Generative AI video platforms.
After a quiet spell, newer players like Kling, Minimax, Runway Gen 3, and Luma Labs took center stage with their headline-grabbing launches. But now, Pika Labs is making a bold comeback with their 1.5 update, featuring explosive new “Pikaeffects”—including ‘squish, inflate, crush, cake-ify, melt, and explode’.
Luma Labs introduces hyperfast video generations (10x faster)
Luma Labs has ushered in a new era of Hyperfast video creation! With 10x faster inference speeds, users can now generate full-quality Dream Machine v1.6 clips in under 20 seconds. In a subtle jab at Runway, Luma Labs emphasised on social media that there are no "turbo" or "distilled" models - just unbeatable quality without compromise, now 10x faster. Available today for all subscribers and API customers.
MiniMax AI is rolling out image-to-video this week
Minimax shook up the Generative AI video space about a month ago with the launch of its text-to-video model. Not only was the model high-quality and free to access -without even requiring a login - but it was built by a team of just six people! This shattered the myth that massive teams and millions (or even billions) of dollars are needed to train video models. However, for professionals, text-to-video is less appealing compared to image-to-video, as the latter offers greater creative control, continuity, and the ability to craft sellable content for clients.
Generative AI Training and Workshops
Our comprehensive training sessions and workshops will massively elevate your ability to create stunning images and so much more - for users at all levels!👇
Kling 1.5 introduces a new lipsync feature
The new “Match mouth type” and “Upload local dubbing” features now allow Kling-generated characters to lip-sync seamlessly with any uploaded audio source. As demonstrated in this clip from X account uncanny_harry, not only do the characters sync their lips impressively, but they also express emotion in the process. Even more remarkable is that the character no longer needs to face the camera for accurate lip-syncing, overcoming a major limitation of previous tools.
California’s controversial AI safety bill vetoed
“California is home to 32 of the world’s 50 leading AI companies,” the governor of California Gavin Newsom said in a statement accompanying his veto of the divisive AI safety bill. “The bill applies stringent standards to even the most basic functions – so long as a large system deploys it. I do not believe this is the best approach to protecting the public from real threats posed by the technology.” Newsom instead stated that they will “develop workable guardrails” that focus “on developing an empirical, science-based trajectory analysis” and work closely with the AI companies in California.
Sign up for next week’s ‘AI Live’ event!
Thursday - 10th October 2024 - 4pm GMT
Each week we host a Live Generative AI podcast on LinkedIn where we talk about the latest AI news as well as answer your burning questions! SIGN UP HERE.
Our favourite post this week!
Each week we feature our favourite post of the week. For more content like this you can follow us on LinkedIn.
Nike Equestrian 2045 | AI Spec Ad
The latest addition to Dave Clark’s renowned AI Spec Ads series is Nike Equestrian 2045, crafted entirely using Magnific AI’s Mystic v2 and Kling AI’s image-to-video model. We enjoyed contributing a remix of Depeche Mode's "Stripped," remixed by Chris Branch ☕️, to bring the project to life 🎵.