Things are moving fast in the tech world. Some big (but anticipated) news came in today. Spurred on by the AI boom, Nvidia has become the world’s most valuable company, overtaking Microsoft. In our last newsletter, we reported that Nvidia overtook Apple to become the second most valuable company.

The 'chips' of the AI game are falling where they may—both metaphorically and literally. And it extends beyond corporate rankings, all the way to regulatory landscapes.

Some important regulatory news from Europe came this week that has likely left Mark Zuckerberg none too pleased. Meta's plans to launch its generative AI models in the European market have hit a bit of a snag. Ireland's Data Protection Commission effectively said, "Nope, not happening." Why, you ask?

It started when Meta sent over 2 billion notices to EU users informing them about its AI training activities. These changes were set to reflect on June 26. But, a nonprofit privacy group called None Of Your Business filed 11 complaints across EU countries, saying Meta isn't following General Data Protection Regulation (GDPR) rules and intends to use years of personal posts, private images, and online tracking data for training its AI.

Meta has basically had a free reign when it comes to using user-generated content to train its AI in places like the US and India, but Europe's strict GDPR rules are causing challenges for Meta and other companies, which are trying to improve their AI capabilities on their apps.

One other key issue around Meta AI is asking users for permission before using their data. While EU users will have that luxury, the rest of us may face more uncertainty regarding how our data is used, with fewer protections and clearer consent processes.

While Meta navigates its latest regulatory battle, the text-to-video AI landscape continues to evolve rapidly. Last week, we extensively spoke about the ‘Sora killer’ from China—Kling AI—in our ‘AI Tutorial Of The Week’ section. Today let’s talk about a new model called ‘Dream Machine’ that hit the market this week.

The model was released by Luma AI and allows new and non-paying users to generate five videos a day for free after they have created an account. Social media claims that it is on par with other AI video generators, so I decided to test it out and gave it a popular worthy-of-a-meme prompt: ‘Will Smith eating spaghetti’.

It took 33 minutes for Dream Machine to generate the following five-second clip and the results were not exactly what I had in mind:

Firstly, the man in the video isn’t Will Smith, and that spaghetti is not going into his mouth. The good news is that Dream Machine is available immediately for use, unlike Kling AI, which is only available for Chinese phone numbers. The bad news is that it’s not very good at following instructions to the T.

But, you can try it for yourself. Head over to our ‘AI Tutorial of the Week’ section for a detailed explanation.

Meanwhile, in Silicon Valley…

The saga between Elon Musk and OpenAI has taken an unexpected turn. After a storm of controversy, Musk decided to drop his lawsuit against OpenAI and its CEO, Sam Altman, bringing a temporary lull in the ongoing drama. Or is it just the calm before the big storm breaks out? With Musk, you never know.

But OpenAI mustn’t lower its guard just yet. A rebellion seems to be brewing among its current and former employees. In an open letter, they claim that the public must be warned about the risks associated with advanced AI and that the companies developing these advanced AI capabilities “have strong financial incentives to avoid effective oversight.”

It seems his own employees are losing faith in Altman’s leadership. Former board member Helen Toner, who led the charge to fire Altman last year, shed some light on the drama of November 2023. She claimed that Altman had kept crucial information under wraps, like the rollout of ChatGPT and his ties to OpenAI's startup fund. She also dropped a bombshell, alleging that two executives had reported psychological abuse by Altman to the board.

But OpenAI is also playing the field to thwart the storm of allegations it has found itself surrounded by. The company roped in retired US Army General Paul Nakasone, who previously led the NSA and the military’s Cyber Command unit, as part of its board of directors.

However, not everyone is happy with Nakasone’s involvement in OpenAI.

Snowden’s views were echoed by the omnipresent Musk, who tweeted, “Can’t wait for OpenAI to have access to my phone …”

Will we witness an Altman ouster 2.0? Probably not. Altman’s come back stronger. He’s got Microsoft in his corner and has secured his spot back on OpenAI's board of directors.

We will keep an eye out as this story develops further. But first, there is more in the AI store.

Imagine you’re a call center operator with a customer on the line who is hurling expletives, which we can't write here to keep things clean. But picture this: There's AI software between you and the rowdy customer that magically transforms their rage into a calm, soothing voice in real-time. Brilliant, isn’t it?

Japan’s SoftBank has developed an AI voice filter to make angry customers sound calm. They did this to curb call center harassment. While the technology does not alter the caller’s words, it significantly modifies the intonation, making it less aggressive.

Dubbed "emotion canceling," this system aims to reduce the stress on call center operators who deal with frustrated customers. You can read all about it here.

And if you have suggestions for next week's coverage (or even in general), you can email me at [email protected]

Did a friend forward this e-mail to you?

IE₊ SUPPORT INTERESTING ENGINEERING
_{Invest In Science And Engineering}

Enjoy exclusive access to the forefront of AI content, highlighting trends and news that shape the future. Join a community passionate about AI, delve into the latest AI breakthroughs, and be informed with our AI-focused weekly premium newsletters. With IE+, AI reporting goes beyond the ordinary - and it is Ad-Free.

NEWS

• Machines on the move: China creates first AI commander for war simulations

  Chinese researchers have reportedly developed a “caged” artificial intelligence (AI) military commander.

• Businessman-backed AI candidate Steve runs for UK parliamentary election

  Steve Endacott, an entrepreneur hailing from southern England, has introduced AI Steve—a groundbreaking artificial intelligence set to contest in the upcoming UK general election.

• AI helps detect Parkinson’s with 100% accuracy 7 years before symptoms

  This breakthrough study highlights biomarkers that may pave the way for earlier drug therapies and offers hope for slowing or preventing Parkinson’s progression.

• Pixelbot 3000: AI-backed Lego printer crafts stunning pixel art mosaics

  The Pixelbot 3000 embodies Lego craftsmanship with modern AI tech, allowing users a streamlined approach to creating intricate brick mosaics based on digital artwork inputs.

MUST-READ

Nvidia overtakes Microsoft to become world’s most valuable company

Riding on the artificial intelligence (AI) wave, chip giant Nvidia has dethroned Microsoft to become the world’s most valuable company.

Nvidia’s share prices rose by 3.5 percent to $135.58 on Tuesday taking its market cap hit $3.335 trillion. Microsoft’s market cap stood at $3.32 trillion, while Apple is the third most valuable company with $3.27 trillion value.

This is the second major rise in the company’s fortunes in under a month. In the first week of June, Nvidia replaced Apple as the world’s second most valuable company.

AI PICTURE OF THE WEEK

The picture in question, which won the People’s Vote and a Jury Award in the artificial intelligence category, has now been disqualified from those categories.

“Nature still outdoes the machine and you helped prove it!” said Astray in a blog post on his website.

TUTORIAL

As promised, here’s a step-by-step tutorial on Luma AI's Dream Machine.

1. Accessing Dream Machine

• Head over to Luma Labs' website.

• You'll see the Dream Machine interface with ‘Try Now’ written at the top right corner of the screen.

• There's a free tier available, allowing you to generate 30 videos per month. If you want access to Dream Machine, you will need to sign up for a free account.

2. Crafting Your Prompt

• After clicking on ‘Try Now,’ you will be transported to another page which will have a text box, wherein you will write a clear and concise description of the video you want to generate.

• Be as detailed as possible. Include elements like objects, scenery, actions, camera angles, and even the desired style (e.g., cartoon, realistic).

3. Using Images (Optional)

• Dream Machine allows you to incorporate an image alongside your text prompt.

• Click the image upload button and select a relevant image that reflects the style or elements you want in the video.

• This can significantly guide the AI towards generating a video that matches your vision.

4. Refining Your Results (Enhance Prompt)

• Dream Machine offers an "Enhance Prompt" button. This is a helpful tool for experimenting with different variations of your prompt.

• Clicking it will suggest alternative phrasings or add details to improve your results potentially.

5. Generating the Video

• Once you're happy with your prompt (and image, if used), click the "Generate" button.

• Dream Machine will process your request and render the video. Depending on the complexity of your prompt, this may take anywhere between a few minutes to an hour.

6. Downloading and Sharing

• Once processing is complete, you can preview the generated video.

• If you like it, you can download the video file.

• Depending on your account tier, you might have limitations on the resolution or length of the video.

That's all for now! Enjoy creating for the rest of the week.