McDonald’s halts AI-driven drive-thru orders due to performance issues.
Hello, AI Enthusiasts!
Our focus today is McDonald’s plans to end its AI-driven automated drive-thru ordering at more than 100 locations, as angry customers put an end to the company’s innovative experiment in the field of AI.
TODAY’S AI News
- McDonald’s is ending its AI drive-thru ordering system
- NVIDIA Sets New AI Training Records
TODAY’S AI WORLD
- OpenAI’s ChatGPT outage： ChatGPT experienced a two-and-a-half-hour outage on June 18, 2024, affecting global users. The disruption impacted various services, including API and Labs. OpenAI quickly addressed the issue, fully restoring service by early morning. This marked the second outage in June, highlighting the importance of robust AI infrastructure.
- GPT-4 outperforms humans in complex tests: GPT-4 can perform better than humans on several standard tests, in terms of how advanced this artificial intelligence system is, these breakthroughs in AI technology will have GPT-4 taking on problems of the highest order of complexity. The current breakthrough would indicate a future in which AI can handle increasingly sophisticated tasks across various industries.
- Apple Enters AI Icon Race： Apple is also in the race to devise an AI icon that should be intuitive and not counter-intuitive to users. This is what an icon Apple has to create across platforms and devices, which means clarity and ease of accessibility for all users. Apple’s design significantly contributes to the general tech community’s effort to balance simplicity and innovativeness.
- Elon Musk’s gamer streams: Elon Musk joined in with one of the most popular streams on the topic of games and even made it more interesting by sharing the updates from Tesla’s latest innovation, talking about the most ambitious projects at SpaceX, and advances in technology linked to brain-computer interfaces from Neuralink. Breaking down complex problems into something any listener can approach, using a casual and engaging style, Musk has drawn a broad audience and started discussions on the future of AI and space exploration.
McDonald’s
McDonald’s pulls AI ordering from drive-thrus
McDonald’s will pause its AI ordering system in drive-thrus after dealing with several accuracy and customer service issues. The technology, which got so much fanfare beforehand, failed to deal with the complex orders and regional accents that just frustrated customers.
The Details
- McDonald’s is canceling a test with IBM for AI drive-thru at over 100 restaurants.
- The technology will be turned off no later than July 26, according to a memo sent to franchisees last week and obtained by CNBC.
- McDonald’s told CNBC that AI would still have a part to play in its drive-throughs in the future.
Conclusion
McDonald’s to stop the test of AI in drive-thrus with IBM after abysmal performance. The system struggled to handle ordering at some drive-thrus where a too-complicated order or heavy customer accent made its processing difficult, forcing a pause to re-evaluate the technology. McDonald’s can now take this feedback into tweaking the technology before future implementation, illustrating just how difficult putting AI into fast-paced environments is. This is a reflection of the challenge of deploying advanced artificial intelligence solutions in real-world, fast-paced environments.
NVIDIA
NVIDIA Sets New AI Training Records
Nvidia has been setting new benchmarks in performance and scale with Hopper GPUs, NVLink, Quantum-2 InfiniBand, and an optimized software stack, among other elements. For example, you are reducing the training time of GPT-3 175B to 3.4 minutes and leading the top MLPerf benchmarks for AI tasks like LLM fine-tuning or text-to-image generation.
The Details
- Performance with Record Setting: NVIDIA set new record performance and scale in MLPerf Training v4.0.
- Advanced Technology: Utilized Hopper GPUs, NVLink, and Quantum-2.
- Optimized Software Stack: Improved performance of AI training with software optimizations.
- Train GPT-3 Efficiency: Speed up the training of the GPT-3 175B model to 3.4 minutes.
- Benchmark Leadership: Performed excellently in AI assignments, including LLM fine-tuning and text-to-image generation.
Conclusion
NVIDIA established new performance records on the latest MLPerf Training v4.0, which enables a new level of generative AI performance. They were using Hopper GPUs, NVLink, and Quantum-2 InfiniBand while on an optimized software stack in this run. Excitingly, results include reducing training time for GPT-3 175B to just 3.4 minutes and leading benchmarks in multiple AI workloads from large language model fine-tuning to text-to-image generation, setting a new level for AI training efficiency and scalability.
AI enthusiasts around the world are eager to learn about your product.
AIExpertReview The latest AI Tools products reviewed by expert. We’ve assisted in promoting Over 500 AI-related Products. Will yours be the next?
What We Can Offer:
- Launch an Advertising Campaign
- Introduce New Product or Features
- Other Business Cooperation
