🎉 Human Brains for Rent, Walmart AI Wins, Upgraded Text-to-Video, Family of Robots, Free Fine-tuning
🎉 Human Brains for Rent, Walmart AI Wins, Upgraded Text-to-Video, Family of Robots, Free Fine-tuning
Rent Brain Cells for $500/mo, Walmart Leads in Commercial Applications of AI, New Text-to-Video Updates from Luma, A New Humanoid Robot Competitor, OpenAI Offers Free Fine-Tuning
Welcome to this week’s edition of AImpulse, a five point summary of the most significant advancements in the world of Artificial Intelligence.
Here’s the pulse on this week’s top stories:
What’s Happening: Swiss startup FinalSpark just launched a service allowing scientists to rent cloud access to "biocomputers" made of human brain cells for $500 a month, in an effort to create AI that uses 100,000x less energy than current systems.
The details:
The system uses organoids (clumps of human brain cells) that can "live" and compute for up to 100 days.
AI models are trained using dopamine for positive reinforcement and electrical signals for negative reinforcement, mimicking natural neural processes.
FinalSpark claims these biocomputers could be up to 100,000 times more efficient for AI training than traditional silicon-based technology.
The organoids and their behaviour are live streamed 24/7, which you can access here.
Why it matters: AI is an energy-hungry industry, and alleviating its dependence on CPUs and GPUs is generally a step in the right direction. That said, using brain organoids for biocomputing is completely uncharted territory and is bound to raise ethical concerns — such as the sci-fi possibility that cell masses somehow achieve consciousness.
What’s Happening: Walmart’s CEO Doug McMillon just reported that the company is using generative AI to increase its productivity, updating 850 million product catalog entries 100 times faster than human-led methods.
The details:
The report came during the company’s Q2 financial earnings call, where McMillon also announced AI improvements to customer search and seller support.
Customers can now use AI-powered search and a new shopping assistant on Walmart’s app and website — it even provides advice for questions like “Which TV is best for watching sports?”.
Walmart is also testing a completely new AI-driven experience for U.S. based marketplace sellers, but the details are not yet available.
McMillon said the company plans to continue experimenting with AI globally across all parts of its business.
Why it matters: Another multibillion dollar company is using AI to increase productivity, but most notably, Walmart is exploring the tech in all areas of its business ops. Whether people should be excited about the endless possibilities ahead or concerned about the relevance of their jobs is a question that’s not going away any time soon.
What’s Happening: Luma Labs just released Dream Machine 1.5, a major upgrade to their current AI video generation model, with higher quality text-to-video, smarter prompt understanding, and better image-to-video capabilities.
The details:
Dream Machine 1.5 builds on the original model’s ability to generate high-quality, realistic 5-second video clips from text and image prompts.
The upgraded model showcases better natural language processing, interpreting and executing prompts at a higher accuracy.
It excels in creating smooth motion, cinematography, and dramatic shots, turning static concepts into dynamic stories, but lags in morphing, movement, and text.
Why it matters: With text-to-image AI generation nearly indistinguishable from reality, the next big frontier is text-to-video — and Dream Machine 1.5 is another big leap forward for realism. While AI video still has some catching up to do, expect fast-moving startups like Luma Labs to close that gap for AI video, fast.
What’s Happening: AGIBOT, a China-based robotics startup, just unveiled a family of five advanced humanoid robots, directly challenging Elon Musk and Tesla’s upcoming Optimus bot.
The details:
AGIBOT’s five new models are both wheeled and biped humanoid robots specifically designed for diverse tasks — from household chores to industrial operations.
The flagship model, Yuanzheng A2, stands 5'9" (175cm), weighs 121 lbs (55kg), and can perform delicate tasks like needle threading.
The company aims to start shipping 300 units by the end of 2024, claiming better commercialization and cost-control abilities than Tesla.
Unitree, another high-performance robot manufacturer from China, also showcased its new G1 mass production-ready robot with better functionality and appearance.
Why it matters: The humanoid robotics and AI race between the US and China is intensifying. While it’s been months since Tesla unveiled its Optimus 2 prototype, four Chinese startups, including AGIBOT revealing five new humanoid robots, have showcased major technical progress in just a few days.
What’s Happening: OpenAI just launched free fine-tuning (up to 1 million tokens per day through September 23) for GPT-4o, allowing developers to customize the model for higher performance and accuracy.
The details:
Developers can now, for the first time ever, fine-tune GPT-4o to improve the model’s structure, tone, and domain-specific instructions for their AI applications.
Fine-tuning is available on all paid usage tiers with training costs of $25 per million tokens, but it is completely free until September 23.
OpenAI suggests that developers should see strong results from fine-tuning with only a few dozen training examples.
Additionally, Google’s Gemini API is giving developers 1.5 billion tokens for free every day on its Gemini 1.5 Flash model and 1.6 million tokens on its Gemini 1.5 Pro model.
Why it matters: Just last week, a company that was granted early access to fine-tune GPT-4o, produced Genie and achieved state-of-the-art scores on both SWE-bench Verified (43.8%) and Full (30.1%) benchmarks. With free fine-tuning now available to all developers, get ready for a new wave of smarter, faster and more capable AI bots.