More hype in the AI Video space from China. Apple talks about AI the Apple way (Apple Intelligence🙄). Amazon using AI for practical purposes. Australian scientists can now detect fire 500 times faster thanks to AI (another practical use case. We love it). Microsoft’s ‘Recall’ feature is probably a bad idea (Who’d have thought 🤔). Nvidia’s CEO is now a rockstar and their AI updates continue to impress beyond belief. Let’s dive in:
Is KLING from China really a Sora Killer?
No. Not necessarily because it isn’t. But because we just don’t know. Not yet, anyway. And we’ve learned not to give into the hype. Having said that, the videos are pretty wild🔥!
🇨🇳OPENAI’S SORA HAS SOME SERIOUS COMPETITION
— Mario Nawfal (@MarioNawfal) June 8, 2024
China’s OpenAI competitor, Kuaishou Technology, just released their latest text-to-video AI tool, Kling.
Kling has the ability to create two-minute videos in crystal-clear 1080p resolution at 30 frames per second, while faithfully… pic.twitter.com/GAgdYkPVzP
We’d only like to call out the fact that if we see anything coming out of China, it’s because they want us to. Not against it, but there’s always more than meets the eye in such a case. Keep an eye out for this.
AI = Apple Intelligence. Really? 🙄
Apple’s WWDC is on 10th June (about 24 hours from when this is likely to be published.) And Apple has done the most Apple thing possible by rebranding AI. Instead of focusing on generative AI features and giving into the hype like everyone else, Apple has decided to focus on features that Apple users are more likely to actually…. use? Siri gets an upgrade. The focus will be on practical uses cases such as reply suggestions, summarization of content and improved privacy. Apple will be partnering with OpenAI (paywall) as well, so it makes sense for them to leave the ChatBots to the bot experts and focus on their user experience. Mark Gurman‘s got the scoop.
Amazon’s Project PI – ‘Private Investigator’ – helps identify defective products
This is the type of AI application that will go down well with the AI doomers as well as the advocates. Using Computer Vision, combined with the use of Multimodal LLMs to identify defects in products or trigger corrective action very quickly.
An example cited in their release blog, was how OCR was used to identify ‘Sell by’ date of dog food which was compared against Amazon’s database to validate that it was OK to ship. If not, it was flagged and fixed before actually being dispatched to the customer. Pretty cool and no where near as ‘hype’ as one would think for AI. Especially given that OCR has been around for a bit. Infinitely more useful than prompting an LLM to generate a picture of a cat on the moon for me. Good going Amazon.
Using AI and space cameras to detected fires in Australia
Yep. You heard that right. Australia is using satellite imagery to detect fires from space so they can be contained in time. The issue with satellite imagery is that small fires are very hard to detect and by the time they’re detectable, they’re really large. For a country with 60,000 annual fires, this can be a problem.
Kudos to the chads over at University of South Australia who have developed a lightweight AI model that can detect fire smoke 500 times faster. The on-board AI does it whilst being 69% more energy efficient at only 16% of the size of it’s ground based processing counterpart. Hyperspectral imagery processing at its finest! The full article is here.
Microsoft’s Recall tool is just as dangerous as it looks
Microsoft has been making all the right moves. For a ‘Boomer’ company, they’ve really shone the way on how to reinvent ones self and stay relevant. Of all the amazing announcements in the last month or so, the upcoming ‘Recall’ feature hasn’t been one of them.
‘Recall’ is basically a feature where Windows takes screenshots of your computer state every few seconds giving you the ability to query it at some point in the future. If this sounds like something out of Black Mirror, it’s because it is. While it sounds cool, it’s a cybersecurity nightmare. Which has been pointed out again and again. And since truth is often stranger than fiction, a hacker tool called ‘Total Recall‘ that neatly stores all your screenshots in 1 place has already been released. Funny, that back when we were ‘script kiddies’, this used to be something we dreamed we could prank our friends in to downloading. Now that Microsoft has decided that it’s a feature and not a bug, I’m not so sure how we feel about it.
Nvidia’s announces Project G-Assistant, overtakes Apple to be the 2nd most valuable company and Jensen is now officially a Rockstar
RIP, Bieber, BTS and One Direction!! 🤣🤣
— Bindu Reddy (@bindureddy) June 5, 2024
Jensen for the win!! pic.twitter.com/ziY8AUI6Qn
We can all agree that Nvidia has been a rockstar for a while now. They’ve absolutely been crushing it in the GPU space. We can add to that their own AI chops that we’ve reported on before. And while they made a some spectacular announcements in the AI space in the past week, we were blown away by Project G-Assist. We won’t do justice explaining it, so just watch this video:
TL:DR: This is Nvidia’s response to Microsoft Copilot’s screen vision. Except it’s built for gamers. Who Nvidia knows very well! Your very own gamer assistant. The assistant not only gives you information, but also optimises the game for you. It auto configures the game settings for FPS, performance, etc…
So it’s not very surprising that Jensen Huang, President of Nvidia, is now a rockstar. The photograph above proves it. Also yes, Nvidia was the 2nd most valuable company by market cap for a bit. Is back to 3rd place, but is likely to regain the 2nd spot. We hope so anyway.
That’s it for this week! If you like this article, please share it with someone who would like to stay up to date with AI news.
