President Trump has stirred debate in the AI world by declaring that anyone who uses the term "artificial intelligence" is the enemy. The comments, made on Truth Social, come as Trump signed the White House Accord on Super Intelligence with six AI companies on September 29. The agreement includes voluntary safety policies such as audits and risk reviews, and a Super Intelligence Force will report on AI risks within 120 days. Several private companies, including Meta and X, have already begun adopting the new terminology.
In a separate development, Mecka AI raised $60 million in a Series A round led by Sequoia Capital. Founded in 2024, the startup collects human motion data by paying people to record everyday tasks like making coffee or folding laundry. These recordings are turned into motion datasets that humanoid robots can learn from. Mecka has already processed over 10 million hours of human motion data in 18 months and plans to expand its contributor network globally.
AT&T is reshaping how it works with law firms through a partnership with OpenAI. General Counsel David McAtee said the company is sending work traditionally handled by junior lawyers back to in-house attorneys, using AI tools like chatbots and agents to access over 100 years of institutional knowledge. McAtee expects the roughly 30 national law firms his team relies on could shrink to 18 to 20 over the next two years.
EasyPark built an AI agent called Parker using Salesforce's Agentforce to identify and qualify B2B leads from existing users. Parker sifts through behavioral data to find high-intent customers, freeing up human sales reps. This marks EasyPark's first production use case for Agentforce. Meanwhile, Arista Networks introduced its Etherlink SU-14 architecture, supporting up to 144 accelerators in a single rack and up to 1,024 XPUs through cross-rack configurations, using Ultra Ethernet Consortium standards.
Research efforts continue across the field. GradSAT addresses gradient domination in Satisfiability Modules Theories solvers using dynamic gradient normalization and a GPU-accelerated PyTorch backend. QRAKEN, a training-free neurosymbolic pipeline, helps large language models query RDF knowledge graphs more accurately, achieving strict F1 scores of 0.643 with GPT-4.1 mini and 0.652 with GPT-5.4 on the CK25 benchmark. Rubrik also expanded its focus, partnering with CrowdStrike for real-time threat detection at its Healthcare Summit 2026, while Fortinet leverages AI to protect AI data centers and prevent employees from using risky AI tools.
Key Takeaways
- President Trump signed the White House Accord on Super Intelligence with six AI companies on September 29, including voluntary safety policies and a task force reporting within 120 days.
- Trump declared on Truth Social that users of the term "artificial intelligence" are the enemy, pushing the term "Super Intelligence" instead; Meta and X are already using it.
- Mecka AI raised $60M in a Series A led by Sequoia Capital to collect human motion data for training humanoid robots, having processed over 10 million hours in 18 months.
- AT&T partnered with OpenAI to use AI tools like chatbots and agents, potentially reducing its law firms from 30 to 18-20 over the next two years.
- EasyPark built AI agent Parker on Salesforce's Agentforce to qualify B2B sales leads, marking its first production use of Agentforce.
- Arista Networks launched Etherlink SU-14, supporting up to 144 accelerators per rack and 1,024 XPUs cross-rack, using Ultra Ethernet Consortium standards.
- GradSAT framework improves constraint solving in SMT solvers through dynamic gradient normalization and GPU-accelerated PyTorch backend.
- QRAKEN pipeline achieved F1 scores of 0.643 (GPT-4.1 mini) and 0.652 (GPT-5.4) on CK25 benchmark, representing 30-32% relative gains.
- Rubrik partnered with CrowdStrike for real-time threat detection and reported 63.95% year-to-date stock gains at $123.73 per share.
- Fortinet uses AI to protect AI data centers and can block employees from using risky AI tools within seconds.
Trump says AI users are the enemy as White House embraces it
President Trump said anyone who uses the term AI is the enemy, even though the White House uses it. He prefers the term Super Intelligence. The comments came during a public appearance at the White House.
White House AI accord and Super Intelligence task force explained
On September 29, President Trump signed the White House Accord on Super Intelligence with six AI companies. The agreement includes voluntary safety policies like audits and risk reviews. A Super Intelligence Force will report on AI risks within 120 days. The accord does not immediately make these commitments into law.
Trump calls AI users enemies in new term war
President Trump warned that anyone calling ChatGPT or Claude by the term Artificial Intelligence will be added to his enemies list. He made the statement on Truth Social. He wants the industry to use Super Intelligence instead. A recent survey showed only 12% of people knew the new term.
Trump separates artificial intelligence from super intelligence
President Trump said the White House is considering how to tell apart artificial intelligence and super intelligence. He also said the White House views anyone using the term Artificial Intelligence as the enemy. The comments were made on Truth Social. The move relates to how advanced AI developers might be regulated.
Trump threatens enemies over AI name change
President Trump said anyone who does not call AI super intelligence will be treated as the enemy. He wants the term Super Intelligence to replace artificial intelligence. He signed an executive order in late September requiring the government to use the new term. Some private companies like Meta and X are already using it.
Mecka AI raises $60M from Sequoia to train robots with human motion data
Mecka AI raised $60 million in a Series A round led by Sequoia Capital. The startup pays people to record everyday tasks like making coffee or folding laundry. It turns these recordings into motion datasets that humanoid robots can learn from. The company has processed over 10 million hours of human motion data in 18 months. Mecka plans to expand its contributor network globally with the new funding.
Robot data startup Mecka AI raises $60M from Sequoia
Mecka AI, founded in 2024, raised a $60 million Series A round from Sequoia. The company collects human motion data to train humanoid robots and other automated systems. It pays people to record tasks like making coffee or fixing cars while wearing body sensors and using smartphones. Mecka aims to do for robotics what data-labeling companies like Scale AI have done for AI language models.
Rubrik stock under scrutiny after security push and CrowdStrike partnership
Rubrik held its Healthcare Summit 2026 on September 30 to highlight its expansion into identity security and AI agent protection. The company partnered with CrowdStrike for real time threat detection. Rubrik shares returned 32.09 percent over 30 days and 63.95 percent year to date at a close of $123.73. Its most followed narrative pegs fair value at about $120.15 per share, suggesting the stock is roughly 3 percent overvalued.
GradSAT framework improves floating-point constraint solving with gradient normalization
GradSAT is a new framework that addresses gradient domination in Satisfiability Modulo Theories solvers. It uses dynamic gradient normalization to balance gradient magnitudes across all clauses at runtime. The system treats each SMT clause as an independent multi-task learning task. GradSAT uses a GPU-accelerated PyTorch backend with symbolic compilation to find high-quality solutions. A bit-precise local search engine then resolves the exact assignment.
QRAKEN pipeline improves natural language queries on knowledge graphs
QRAKEN is a training-free neurosymbolic pipeline that helps large language models query RDF knowledge graphs accurately. It produces TTQL, a compact description of multi-hop patterns and conditional frequencies. Online, TTQL guides the language model while deterministic checks provide diagnostics for refinement. On the CK25 benchmark, QRAKEN achieved strict F1 scores of 0.643 with GPT-4.1 mini and 0.652 with GPT-5.4. These results represent relative gains of 30 percent and 32 percent over the strongest recomputed participant.
9th Annual State of AI Report Released for 2026
The author released the 9th annual State of AI Report after months of research. Their interest in AI began with a PhD in cancer research and computational biology in the early 2010s. Since 2013, they have invested in companies building and applying AI to accelerate technological progress. The report aims to share what they are learning and help more people understand AI.
Arista Launches New Rack-Scale Ethernet Solutions for AI
Arista Networks introduced its Etherlink SU-14 architecture, which supports up to 144 accelerators in a single rack and up to 1,024 XPUs through cross-rack configurations. The platform uses Ultra Ethernet Consortium standards and features like Dynamic Load Balancing to improve AI workload performance. It also integrates liquid-cooled 7060EX7 switches with up to 102.4 Tbps of capacity per switch. Next-generation optics can raise networking capacity from 1.6 Pb/s to 6.5 Pb/s per rack.
EasyPark Builds AI Agent Parker to Qualify Sales Leads
EasyPark built an AI agent called Parker based on Salesforce's Agentforce to identify and qualify B2B leads from existing users. Parker sifts through behavioral data to find high-intent leads and free up human sales reps. EasyPark's ANZ country director Jason Marks said the company has grown from a handful of staff four years ago to over 60 in Australia and 4500 globally. Parker is EasyPark's first production use case for Agentforce.
AT&T Uses AI to Reduce Reliance on Law Firms
AT&T General Counsel David McAtee has stopped paying law firms to do many assignments traditionally handled by junior lawyers, sending the work to AT&T lawyers instead. The company formed a partnership with OpenAI to make over 100 years of institutional knowledge easier to access through AI tools like chatbots and agents. McAtee said the 30 or so national law firms his team relies on could shrink to 18 to 20 firms over the next two years. AT&T lawyers have done thousands of hours of work this year that law firms would have traditionally done.
Fortinet AI Cybersecurity Stock to Buy and Hold
Fortinet's physical cybersecurity components and software are vital resources for AI data centers. The rising use of agentic AI increases the need for cybersecurity, especially as physical AI becomes mainstream. Fortinet beat its second-quarter guidance and raised its full-year revenue outlook. The company uses AI to fight AI and can prevent employees from using risky AI tools within seconds.
Sources
- Trump: Anyone saying "AI" is "THE ENEMY!" The White House uses it
- White House AI Responsibility Accord & Super Intelligence Task Force: What Businesses Need to Know
- Trump declares anyone who uses the term Artificial Intelligence to be ‘The Enemy!’
- Trump draws a crucial line between ‘artificial intelligence’ and ‘super intelligence’
- Trump says anyone who does not call AI 'super intelligence' is 'the enemy'
- Mecka AI Lands $60M Sequoia Round for Robot Training Data
- Robot data startup Mecka AI nabs $60M from Sequoia
- Is Rubrik (RBRK) Overvalued After Its Security Push And CrowdStrike Partnership?
- Accelerating Floating-Point Satisfiability Solving via Gradient Normalization
- Natural Language Questions as an Interface for Knowledge Graphs: QRAKEN Graph Distillation and Semantic Self-Healing
- The State of AI Report 2026
- Can Arista's New Rack-Scale Ethernet Solutions for AI Stoke Growth?
- EasyPark builds an AI agent to help it qualify B2B sales leads
- How AT&T is using AI to reduce its reliance on law firms
- 1 Millionaire-Maker AI Cybersecurity Stock to Buy and Hold
Comments
Please log in to post a comment.