- thereach.ai
- Posts
- Nvidia's CEO Thinks We Need Only 2 Trillions, Not 7, Apple Takes On Midjourney
Nvidia's CEO Thinks We Need Only 2 Trillions, Not 7, Apple Takes On Midjourney
Sorin Ciornei
February 13, 2024
Welcome, AI enthusiasts
Nvidia’s CEO thinks we only need perhaps 2 trillions instead of 7 for AI, Gen Z is ready for the AI age, a new AI will join the race and Apple will take on Midjourney
In today’s newsletter:
Gen Zers are comfy with AI, for now
Windows will add Super Resolution to Windows, powered by AI
ElevenLabs let's you monetize your voice
A new startup joins the AI race - Reka
Apple takes on Midjourney - a new AI model for instruction-based image editing.
NEWS OF THE DAY
Sam Altman Vs. Jensen Huang —
Nvidia CEO Aims for Efficiency
What: Nvidia CEO Jensen Huang anticipates that ongoing computing advancements will limit AI costs significantly below the reported $7 trillion target set by Sam Altman for semiconductor production. This statement made waves during a discussion at the World Government Summit in Dubai.
Key points:
Huang foresees that the acceleration in chip speed will enhance efficiency, resulting in a substantial reduction in overall AI expenditures, well below Altman's proposed $7 trillion chip initiative.
The CEO strongly advocated for nations to leverage cultural data in constructing "sovereign AI" infrastructure and intelligence, underscoring the importance of countries having ownership of their data.
Furthermore, he projected a doubling of the global cost of AI data centers to reach $2 trillion within the next 5 years.
In another noteworthy development, Nvidia made waves in the stock market as its market value soared to $1.8 trillion, surpassing Amazon on Monday.
Noteworthy: We should also remember Sam Altman recently hired a new Hardware VP. Is he going the same path as Google and FB with building their own infrastructure and consolidate the ecosystem?
EDITORIAL
What: Artificial intelligence (AI) has become an integral part of our lives, revolutionizing how we work and interact with technology. In this dynamic landscape, three prominent AI chatbots—ChatGPT, Microsoft Copilot, and Gemini—vie for supremacy. Let’s look into their nuances to help you make an informed choice.
Key points:
💬 ChatGPT, Microsoft Copilot, and Gemini are prominent AI chatbots vying for supremacy.
🧠 AI chatbots like ChatGPT, Copilot, and Gemini are evaluated based on their ability to write code, generate creative writing, and solve riddles.
📊 Gemini, Google's AI chatbot, offers speedy responses, accurate answers, and full Google integration.
🌐 Microsoft Copilot provides up-to-date information, access to GPT-4, and image creation capabilities.
🛠 ChatGPT offers a simple interface, frequent updates, and various features, though some users have reported degraded performance over time.
Noteworthy: The question of ‘whatever works for you’ might be the wrong one – a better question is which ecosystem you are more comfortable using or perhaps which one you are already locked into. In the end, these AI solutions will likely provide very similar results across the board, both today and in the future.
LATEST NEWS
💡 Gen Z embraces generative AI (GAI) at school and work.
🎓 "AI native" Gen Zers see AI skills as a career advantage.
📚 College students are enrolling in GAI courses for a job market edge.
🚀 A significant portion of Gen Z plans to use GAI in their careers.
Read the full study here.
ElevenLabs now let’s you monetize your voice
Calling all Voice Actors.
Now you can share your voice in the ElevenLabs Voice Library and earn when it's used.
With as little as 30 minutes of audio, you can create a voice that’s ready to be used in projects around the world.
— ElevenLabs (@elevenlabsio)
5:00 PM • Feb 12, 2024
Now you can share your voice in the ElevenLabs Voice Library and earn when it's used. With as little as 30 minutes of audio, you can create a voice that’s ready to be used in projects around the world. You can also read the deep dive we did of ElevenLabs last year.
🎮 Microsoft is testing an "Automatic Super Resolution" AI-upscaling feature for gamers on Windows 11.
🆙 Upscaling enhances video and image quality of supported games while improving performance.
🧠 Newer techniques use AI and deep learning algorithms to predict and add details, maintaining quality.
🖼️ Apple released the open-source AI model "MGIE" for instruction-based image editing.
🧠 MGIE leverages multimodal large language models (MLLMs) to interpret natural language instructions for pixel-level manipulations.
🌐 MGIE integrates MLLMs to derive instructions and generate visual imagination, optimizing the editing process.
AI startup Reka has unveiled an impressive 21 billion parameter model, trained entirely from scratch. Distinguishing itself as natively multimodal, it outperforms benchmarks set by Llama 70B and Mixtral across various key metrics. Notably, aligning with the founders' previous affiliations, it also rivals the capabilities of Gemini Pro. The model is now accessible for use in Reka's playground.
MIDJOURNEY PICK OF THE DAY
photography in the style of Yinka Ilori of an Ikea showroom
Check out our guides on how to generate the best Midjourney pictures, from fashion, nature, street art or illustration.