Welcome to another edition of Horizon AI,

In a wide-ranging conversation with Lex Fridman, Sam Altman discusses many topics including GPT-5, AGI, and Elon Musk’s lawsuit. Meanwhile, Nvidia just presented what they claim is the ‘world’s most powerful chip.’

Let's jump into it!

Read Time: 4.0 min

Here's what's new today in the Horizon AI

• Sam Altman Reflects on OpenAI and Shares Thoughts on The Future of AI 🎙️

• Nvidia Unveils Blackwell B200 - The World's Most Powerful AI Chip

• AI Research: Google Unveils Groundbreaking AI System to Animate Still Photos

• AI Tutorial: Turn your pictures into stickers in seconds

• AI Tools to check out

• The Latest in AI and Tech 💡

AI News

OPENAI

Sam Altman Reflects on OpenAI and Shares Thoughts on The Future of AI 🎙️

Sam Altman gave a two-hour-long interview to podcaster Lex Fridman, in which they discussed a wide array of topics.

Here are the main takeaways from the interview:

• Sam Altman opened up about the OpenAI board saga, describing it as the most painful professional experience of his life, but one that ultimately facilitated his personal growth.

• The CEO also addressed the ongoing legal dispute with Elon Musk, expressing sadness but confidence in overcoming the disagreement.

• He confirmed the arrival of 'an amazing new model this year' but wouldn’t commit to it being called GPT-5.

• He also hyped up the capabilities of GPT-5 by stating that GPT-4 'kinda sucks' and suggesting that in the near future, we'll view it as we currently do with GPT-3.

• Altman suggests that compute will become the currency of the future. He envisions a future where compute is the most precious commodity.

The interview also covers topics such as OpenAI’s new Sora, AGI, and the collective responsibility of AI, among others. It is definitely worth checking out if you have the time.

You can listen to the full podcast here or read the transcript on Lex Fridman’s website.

TOGETHER WITH MIT

AI: Balancing Risk and Return

Innovative technologies are revolutionizing business as we know it, and they’re more accessible than ever. But to truly harness the transformative potential of AI, you need to know how and when to use it. And which pitfalls to avoid.

The six-week Artificial Intelligence: Implications for Business Strategy online short course from MIT Sloan School of Management and MIT Computer Science and Artificial Intelligence Laboratory explores AI’s business applications and challenges. 

Choose this program to:

• Optimize your business: Leverage AI, ML, and robotics to drive efficiencies, improve productivity, and support your growth.

• Develop a strategic roadmap: Apply your knowledge to effectively integrate AI into your business.

• Gain a dual perspective: Benefit from a course designed by two prestigious schools — the MIT Sloan School of Management and the MIT CSAIL.

• Conveniently build career-critical skills: Follow a program that fits your schedule and benefit from 24/7 support and various payment options.

NVIDIA

Nvidia Unveils Blackwell B200 - The World's Most Powerful AI Chip

Nvidia has revealed the Blackwell B200 GPU, dubbed the "world's most powerful chip" for AI. This next-generation processor promises to revolutionize the AI industry with unprecedented performance capabilities.

Key highlights include:

• 4x faster training performance compared to Nvidia's flagship H100 chip

• Up to 25x greater energy efficiency

• 30x inference performance for complex models like GPT-4

• Even older models like GPT-3 will experience an 8x speedup

Nvidia CEO Jensen Huang declared that the Blackwell platform will drive the 'new industrial revolution' by enabling the training of large language models scaling up to 10 trillion parameters (as a reference, GPT-4 contains just 1.8T parameters).

The performance gains are attributed to Nvidia's new generation Transformer engine and an enhanced NVLink interconnect. This technological leap positions Blackwell as the prime catalyst for pushing the boundaries of what's possible with AI.

Source

AI RESEARCH

Google Unveils Groundbreaking AI System to Animate Still Photos

Researchers at Google have developed VLOGGER, an AI system capable of generating strikingly lifelike videos from just a single still image and audio clip.

Details:

• VLOGGER can take a photo of a person and an audio clip as input, and output a video showing that person speaking the words while making natural facial expressions, head movements, and hand gestures synchronized to the audio.

• The model was trained on a massive dataset containing over 800,000 diverse identities and 2,200 hours of video data, allowing it to learn rich patterns of human behavior.

• It could be applied in presentations, education, online communication, and human-computer interaction interfaces.

While not yet perfect, VLOGGER represents a significant leap in animating still images. We may soon live in a world where it becomes increasingly difficult to discern computer-generated videos from reality.

Source

AI Tutorial

Turn your pictures into stickers in seconds.

You can now effortlessly turn any image into a sticker for free. Just follow these steps:

1. Go to Replicate’s face-to-sticker

2. Upload your picture

3. Add a prompt and change the settings

The prompt should describe your picture, you can also change the settings like width, height, prompt strength, upscaling, etc.

4. Generate the sticker

Click on the “Run” button and wait a few seconds for your image to turn into a sticker.

Examples:

AI Tools to check out

🎨 Musavir: User friendly image generator that creates realistic and stylized images, hyper-localized images, and personalized avatars.

🎧 iListen: Summarise any article or webpage into concise, easy-to-digest podcasts.

🤖 Bytebot: Enhance and simplify your browser automation using AI-enabled SDK.

🎉 CelebrateAlly AI: Save time and create unforgettable celebrations. Craft heartfelt speeches, design personalized parties, search for the perfect gifts, and more.

📅 Dola: Powerful AI calendar assistant managing your personal calendar and group calendar by just chatting.

The latest in AI and Tech

YouTube introduces a new AI content labeling tool for transparency

YouTube has introduced a new self-labeling tool for creators, allowing them to disclose when their videos contain AI-generated or synthetic material. Appearing during the upload and posting process, the checkbox requires creators to identify "altered or synthetic" content that appears realistic.

Roblox unveils two new AI features to accelerate 3D Creation

Roblox has unveiled two AI technologies aimed at streamlining the content creation process on its platform. These AI-driven features significantly reduce the time required for creating avatars and 3D models.

FTC inquires about Reddit's AI deals

The United States Federal Trade Commission is investigating social media platform Reddit over its data licensing practices for AI training. This investigation comes as Reddit prepares for its long-awaited initial public offering.

SEC fines two advisors, alleging they lied about AI capabilities

The U.S. Securities and Exchange Commission (SEC) has taken action against two investment advisory firms, Global Predictions and Delphia, accusing them of engaging in "AI washing" – making misleading claims about their use of AI in managing investments.

Homeland Security rolls out AI pilot projects

The Department of Homeland Security is intensifying its exploration of generative AI with the announcement of three new pilot projects on Monday. Complementing these pilots, DHS held a recruiting event in Silicon Valley last month, aiming to onboard at least 50 AI experts into its newly formed "AI Corps."

That’s a wrap!

Thanks for sticking with us to the end! Let’s stay connected on LinkedIn and Twitter.

Not subscribed yet? Sign up here and send it to a colleague or friend!

See you in our next edition!

Gina 👩🏻‍💻