Notes I took from the artificial intelligence news that caught my eye between February 16 and February 20.

  • I ended the last episode with Google’s Gemini 1.5 announcement. Three hours after I wrote that post, a bombshell came from OpenAI, and they announced a new artificial intelligence model called Sora. With this new text-to-video AI model, it can convert text into video for up to 60 seconds. And the videos coming out of the first version of Sora are truly eye-catching. For now, end-users will have to wait a little longer for the video AI model, which is currently being tested by its internal ‘Red Team’ and some selected artists. If you go to the official page where the technical explanation is made above, you can see how good the current video outputs are. The fact that artificial intelligence, which has just entered our lives, is so impressive even in its infancy stages both scares and excites us for the coming years.

Introducing Sora, our text-to-video model.

Sora can create videos of up to 60 seconds featuring highly detailed scenes, complex camera motion, and multiple characters with vibrant emotions. https://t.co/7j2JN27M3W

Prompt: “Beautiful, snowy… pic.twitter.com/ruTEWn87vf — OpenAI (@OpenAI) February 15, 2024

  • Along with Gemini 1.5, the Pro version was also released. Google is sharing videos about the features of its new ‘pro’ artificial intelligence model. I couldn’t believe the Gemini 1.5 Pro version’s ability to understand the plot and analyze even the finest details in the video below that I saw on Twitter . There isn’t much to say. Just watch.
  • Nvidia, at the center of this whole artificial intelligence phenomenon, has opened the doors of its supercomputer Eos to the world. The supercomputer containing 4608 H100 GPUs, which it uses for generative AI, is the newest enterprise-focused supercomputer designed for advanced AI development at a data center scale, and the company’s fastest AI supercomputer. With its processing power, Eos is the ninth most powerful supercomputer in the world. It is said that a single H100 GPU costs between 30-40 thousand dollars. You do the math on the size of the investment made.
  • It seems that the demand for hard disk drives (HDD) is recovering after a significant decline recorded in the last two years. The fact that SSDs are more preferred by end-users might be confusing, but on a very large scale, HDDs are still the most economical and logical solution. Reuters reported last Friday that Sony’s semiconductor unit is preparing to start mass production of hard disks due to increasing AI demand. They have agreed with Seagate for mass production. Let’s hope this doesn’t reflect negatively on us end-users in terms of price.
  • Apple will integrate an AI tool into Xcode this year that will make it easier for developers to write code, according to reports . I am very curious about how Apple will integrate AI into its products like the iPhone, Mac, iPad, and Watch later this year. Because the gap continues to widen. Although Apple brought some features to iPhones later, it always managed to steal hearts with its user-friendly structure. Let’s see what will happen this time.
  • Google, or rather Alphabet CEO Sundar Pichai, gave an interview and revealed that he uses 20 smartphones at the same time. The part that interests us is what he said about artificial intelligence. In the interview, he said that AI is the most important technology produced by humanity. So much so that he compared AI to the discovery of fire and electricity.
  • According to Bloomberg , Reddit has signed a deal “worth about $60 million annually” to allow content on the site to be used to train an AI model. According to the news, which did not name the AI company, I understand that announcing such a deal ahead of Reddit’s IPO at a $5 billion valuation has commercial and strategic importance. We will see how this agreement will be received by the users.
  • Eliezer Yudkowsky, an AI researcher famous for his pessimism, made a new prediction about the future of humanity. According to our friend’s feelings, we might be 10 years or 5 years, or even 2 years away from a Terminator or Matrix-like dystopian threshold. Fortunately, we won’t have to wait too long to find out if his feelings turn out to be true.
  • Meizu, one of China’s leading technology companies known for its smartphones, recently made a major strategic shift. The company has decided to stop developing new smartphones and will focus entirely on artificial intelligence with all its resources. The company’s plans reportedly include releasing a new mobile operating system suited for the AI era in 2024, and launching its first AI hardware product later this year. Such a radical decision requires courage. I am eagerly waiting to see what Meizu will come up with.
  • Last week, I wrote about Sam Altman’s dream of producing $7 trillion worth of AI chips. With the rise of artificial intelligence, it’s not just OpenAI that is trying to keep Nvidia from dominating the playing field alone. Now, Japan’s Softbank is preparing for a $100 billion investment to produce AI chips. According to a report by Tomshardware , it looks like $30 billion will come from Softbank itself, and they will knock on the doors of some wealthy Arab countries for the remaining $70 billion. While everyone is trying to board the AI train from somewhere, it looks like the next 5 years are going to be very tough.
  • I can’t share many AI news stories in the medical field here because that is a whole different ocean. The research is endless. But today I will mention a piece of news that caught my eye and could be very useful in daily life. Researchers at MIT have developed a new AI model for drugs that shouldn’t be taken together. Using a machine learning algorithm, this model can predict interactions that might interfere with a drug’s effectiveness.
  • While we are in the medical field, let me mention another interesting study. Researchers at Stanford University School of Medicine have developed a new method that uses artificial intelligence (AI) to analyze brain scans. This method managed to reveal previously unknown differences between male and female brains. Differences were observed in the connections between different brain regions and in the brain’s wave activity. The model can distinguish between male and female brain scans with over 90% accuracy. Among the differences found:
    • In male brains, the connections linking the two hemispheres of the brain were stronger.
    • In female brains, there were more connections between different regions of the brain.
    • Male brains tended to produce slower brain waves known as alpha waves.
    • Female brains tended to produce faster brain waves known as beta waves.
  • The advancement of artificial intelligence technology and tools at a pace that very few can comprehend, especially in terms of applications and possibilities, is emerging as a security issue at all levels and an area of liability for companies. In this context, NVIDIA has joined the US Artificial Intelligence Safety Institute Consortium (AISIC) for safe and responsible AI. The consortium, created by the US government earlier this month, currently has more than 200 members .
  • Let’s end this episode with a news piece from ElevenLabs, which is connected to OpenAI’s Sora video AI mentioned at the beginning of the article. ElevenLabs announced its AI technology that adds sound effects to videos. It demonstrated this on its own blog by adding audio to the videos used in Sora’s presentation. Thus, it reminded us of the question we are curious to answer: When will a movie completely prepared by artificial intelligence—without humans, cameras, microphones, or screenwriters—come to theaters?

We were blown away by the Sora announcement but felt it needed something…

What if you could describe a sound and generate it with AI? pic.twitter.com/HcUxQ7Wndg — ElevenLabs (@elevenlabsio) February 18, 2024