OpenAI is advancing ChatGPT on multiple fronts. GPT-5.6 Sol is becoming more capable, free users now have unlimited text conversations, and a new reasoning slider lets users adjust the modelâs level of reasoning depending on the complexity of the task @grok
The most interesting development, however, isnât the interface. OpenAI says GPT-5.6 Sol reduced responses containing at least one factual error by 68% compared with GPT-5.5 Instant on financial, medical and legal prompts.
This suggests that the next stage of AI competition wonât be defined solely by speed or model size. It will increasingly be about reliability, factual accuracy, source attribution, and the ability to correctly handle dates, numbers, rules and assumptions
Plus and Pro users also now have a slider to choose how much reasoning effort ChatGPT puts into each response.
We think itâs easier to use, and weâre listening to your feedback.
The new GPT-5.6 Sol powers all chats for paid users, including Instant, creating one consistent experience.
In our high-stakes factuality evaluation covering finance, medicine and law, the new GPTâ5.6 Sol produced 68% fewer responses with factual errors than GPTâ5.5 Instant.
The AI race is no longer just about models. Itâs becoming a race to control the entire technology stack.
It began with NVIDIA @grok whose GPUs became the foundation of the generative AI revolution.
Then came TerraFab, highlighting the next phase of AI infrastructure by focusing on large-scale computing capacity and the industrial foundations required to power future AI systems.
Companies soon realized that relying exclusively on third-party hardware would eventually become a strategic limitation.
OpenAI introduced its custom AI chip project, Jalapeño.
Meta launched its Iris initiative.
Microsoft developed Maia.
Google continued investing in its TPUs.
Amazon expanded with Trainium and Inferentia.
Now Anthropic is assembling a dedicated engineering team to design custom chips optimized for Claude.
This isnât about replacing NVIDIA overnight.
Itâs about diversification, optimization and strategic independence.
The future belongs to companies capable of co-designing hardware and AI models as one integrated system.
The next generation of AI wonât be defined solely by the most powerful model.
It will be defined by those who master the entire stack:
âą Silicon
âą High-bandwidth memory (HBM)
âą AI accelerators
âą Data centers
âą AI models
âą Cloud infrastructure
âą Software
âą Applications
AI is no longer just a software industry. It is becoming a fully integrated industrial ecosystem.
The winners wonât simply build better models. They will build the infrastructure that powers the AI economy
NEW: "Manufacturing is exploding at the fastest pace in many years. Nobody thought this would ever happen again."
President Trump highlights a surge in U.S. manufacturing, declaring that factory growth is expanding at a rapid pace in a public address in Las Vegas.
On Aug. 12, you have the chance to experience a total solar eclipse from anywhere in the world.
Starting at 1:15pm ET (1715 UTC), we'll stream live as the Moon passes in front of the Sun, bringing the views and the science straight to you.
Get safe viewing tips in this video.
@imagine I think Grok Imagine has potential far beyond creative content. Imagine an AI fashion stylist that analyzes your body shape, style preferences, and budget, then creates personalized outfits from Zalandoâs catalogâall with realistic virtual try-ons before you buy
@imagine Grok Imagine has huge potential beyond content creation. I could see it evolving into medical visualization, helping plastic surgeons and dentists simulate treatment outcomes for patients before procedures, much like todayâs 3D imagingâonly faster and more interactive