The week commenced with a flurry of activity on X (perhaps we can drop the "formerly called Twitter" tag at this stage). This time, the spotlight fell on Yann LeCun, Meta's chief AI scientist and a prominent figure in the AI community, engaging in a verbal clash with Tesla lord Elon Musk.

LeCun, who has criticized the owner of X in the past for his ‘vengeful politics,’ was replying to a post by Musk who was calling on his followers to join his startup xAI. LeCun remarked, “Join xAI if you can stand a boss who… claims to want a "maximally rigorous pursuit of the truth" but spews crazy-ass conspiracy theories on his own social platform.”

Oof! Is it just me or is it getting hot in here? And there’s a lot more to where that came from. You can read the rest of their exchange below:

Apart from the tug-of-war over retaining the best of a small and elite pool of AI talent, Musk and xAI had a great week, raising $6 billion in a new funding round to bolster their capital and stay competitive against rivals such as OpenAI, Microsoft, and Google.

ChatGPT outage amid OpenAI whistleblower allegations

Also, ChatGPT was down for roughly two hours yesterday with a select few users unable to get a response to their prompts. OpenAI fixed it soon thereafter. But more problems mount for the frontrunner AI company as nine whistleblowers have said that the company hasn’t done enough to prevent their AI systems from becoming dangerous, reported the New York Times.

This comes after former OpenAI board member Helen Toner revealed some murky details about the mysterious and unceremonious firing of Sam Altman last year, which we talked about in our previous newsletter.

In troubling news for journalists, a spate of news organizations have inked deals with OpenAI to make it a more credible source of information. The latest to partner with the creator of ChatGPT is Vox Media.

This follows the news of OpenAI’s collaborations with The Atlantic, Associated Press, and Semafor. Why is this troubling? The industry worries it could lead to misinformation on news platforms and even possible job cuts.

In other news…

ElevenLabs released a generative-AI tool called ‘Sound Effects’ that can create ‘any sound imaginable’ using text prompts. More on that in the upcoming ‘Tutorial of the Week’ section.

Software developers rejoice… while they can. French AI startup Mistral has introduced ‘Codestral,’ a programming-focused AI model. With a staggering 22 billion parameters, Codestral is designed to tackle coding tasks across more than 80 programming languages, surpassing its predecessors, claims the company. It supports common programming languages like Python, C, C++, SQL, Java, C, C++, Swift and Fortran.

Venturing further eastward, India’s elections, shadowed by the pervasive threat of highly convincing AI-generated deepfakes, announced their results today. Despite their promises, the incumbent government of Bharatiya Janata Party (BJP) fell short of their expected tally. It's worth noting that the results were not AI-generated, although the BJP might have preferred otherwise.

What can you expect in the next edition of AI Logs?

We will be looking at Apple’s offerings at the Worldwide Developers Conference which will be held from June 10 to 14. Tons of AI related announcements are expected from the iPhone maker, with CEO Tim Cook consistently emphasizing that significant waves of AI advancements are slated to arrive on Apple devices throughout 2024. Leaks are surfacing about the details of a probable iOS 18 launch. We’ll be on the lookout!

While Apple has been playing it cool with AI launches compared to its rivals, all eyes are now eagerly awaiting their next move, especially after Google and OpenAI dropped the bomb with their more advanced Siri-like voice assistant AI models—Project Astra and GPT-4o, respectively.

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

• World’s 1st AR laptop with 100-inch workspace that fits in your backpack

  Sightful has unveiled Spacetop G1, an innovative laptop that marries augmented reality (AR) with day-to-day functionality.

• Raspberry Pi 5 gets a powerful AI brain upgrade from Israeli firm

  Raspberry Pi has partnered with chipmaker Hailo to bring artificial intelligence (AI) capabilities to its popular board-based computing systems.
  

• US roots for AI in warfare, awards Palantir contract for Maven prototype

  The US Department of Defense has awarded Palantir a $480 million contract to begin building the prototype of the Army’s Maven Smart System project.
  

• China: Robot doctors at world’s 1st AI hospital can treat 3,000 a day

  A Chinese state media outlet has reported that the country has developed its first AI hospital town, a concept in which virtual patients are attended to by AI doctors.

MUST READ

AMD challenges Nvidia with ‘world’s fastest PC processors,’ new AI chips

AMD has introduced its newest AI processor, the MI325X. This advanced version of the MI300 series is expected to be available in the fourth quarter.

The releases and announcements from AMD came just a day after Nvidia released its next-gen AI chip, ‘Rubin.’

AI PICTURE OF THE WEEK

Over the week, the ‘All Eyes On Rafah’ Instagram template became the most shared and viral AI-generated image with over 50 million shares on the platform and millions more on other social media platforms. The image was used as a means to draw attention to the situation currently unfolding in the southernmost city of the Gaza strip—Rafah—where hundreds and thousands of displaced people have taken shelter from heavy firing.

The picture above is based on an AI-generated image first shared by Zila Abka on Facebook in February. She made it using Microsoft Image Creator. It was recently that an Instagram user turned it into a template using an AI image generator tool.

TUTORIAL

Using ‘Sound Effects’, the latest AI offering on the market

Last week, ElevenLabs announced a text-to-speech AI tool called ‘Sound Effects’ that can generate sounds of pretty much anything (think: police sirens and firework explosions) from a textual prompt.

I gave it the prompt ‘create a sound for a building being constructed’ and the AI tool gave me four different sound options. The first three sounded like a strong metal getting banged against a wall, while the last sounded like heavy gunfire. One can also give fun prompts like ‘An elderly woman making sounds while removing her dentures.’

Below is a step-by-step process for how to operate Sound Effects:

Step 1: Access the Tool

ElevenLabs offers a free tier with limited generation options. To use it, head over to their website and click on "Sound Effects".

Step 2: Describe Your Sound

In the text box provided, type a clear and concise description of the sound effect you want to generate.

Be as specific as possible. You can include details like:

• Type of sound (footsteps, doorbell, thunderstorm)

• Additional details (boots vs sneakers, heavy rain vs drizzle)

• Environment (forest, stone hallway)

Tip: The character limit is 300 characters, so be descriptive but concise.

Step 3: Generate

Once you're happy with your description, click the "Generate Sound Effects" button.

Step 4: Review and Download

• ElevenLabs will generate four different audio samples based on your prompt.

• You can listen to each sample by clicking the play button next to it.

• Choose the sound effect that best suits your needs.

• Click the download icon next to your chosen sample to save it to your device.

You can also use the tool to generate short musical pieces or character voices using similar techniques. For more advanced usage and unlimited generation, you can consider their paid subscription plans.