Hi Again,
Happy Labor Day!
Welcome to September- historically worst month for the stock market! Probably we will not have a significant correction. Why? When most people expect the market to go one way, it goes in the opposite direction. Tom Lee, who has been correct about the market for years believe that September will be a good month for the market. We also have about 6 Trillion in money market funds so as soon as we get a correction, we could see a v shaped recovery. Nothing much to report. As Jim Cramer said on 8/31/26, “August was a nothing month”. About 60% of analysts believe that the Feds will increase rates in September. I do not agree. It is not appropriate to increase rates 6 weeks prior to the mid term elections. However, unless trends change, Feds might raise rates in 2026. Global bond rates are rising due to huge government deficits in most countries.
Bob Lang is one of the country's top options traders, an expert market technician, and a highly sought-after mentor and teacher. On 8/31/26, he stated that looking at copper prices, it is clear that the US economy is on fire. 2023 to 2026 were the best years for the copper market. Take a look at the stock for Hud Bay Minerals (HBM). Prior to 2024, it could not go over $5. Now it is at $27.50 with a reasonable PE (price to earnings) of 16.83.
Now for the AI saga. On September 3, 2026, a major, unprecedented simultaneous outage knocked down the world's leading AI chatbots—including OpenAI's ChatGPT, Anthropic's Claude, and xAI's Grok. While each platform initially offered distinct official explanations for their individual downtime, reports tracked the shared culprit to a regional failure inside Microsoft Azure's cloud infrastructure.
Even though these AI platforms are operated by fierce competitors, they heavily rely on overlapping cloud hosting and network routing backbones. The Azure Connection: Tech analysts confirmed that a breakdown in Microsoft Azure's East US region served as the core bottleneck. Because these models lean on Azure for massive compute clusters, cross-cloud peering, or API traffic routing, a failure at the cloud layer simultaneously crippled multiple platforms. How Gemini Survived: Notably, Google's Gemini avoided the blackouts. It remained fully operational because it operates entirely on Google's independent, in-house infrastructure, using its own custom TPU clusters and Google Cloud Platform (GCP). The outages lasted roughly 90 minutes to a few hours on Thursday morning before services were successfully mitigated and restored.
As with other US industries, China ahs been stealing US intellectual property to create their AI versions. Chinese efforts to acquire U.S. artificial intelligence technologies have intensified, primarily taking the form of high-profile insider theft and large-scale digital "distillation" campaigns. U.S. intelligence, government officials, and leading technology firms have repeatedly warned that Chinese entities are utilizing both covert intellectual property exfiltration and algorithmic replication to close the technological gap between the two nations. The most prominent case of direct insider intellectual property theft involves Linwei Ding (also known as Leon Ding), a former Google software engineer. The Theft: Ding was found to have systematically uploaded more than 500 confidential files containing Google's AI trade secrets to his personal cloud account. The stolen files detailed the hardware infrastructure, custom Tensor Processing Unit (TPU) chips, and software platforms used to power Google’s supercomputing data centers. The Chinese Connection: While employed by Google, Ding secretly aligned himself with China-based AI startups and even spent months in China seeking funding to launch his own AI company using the stolen tech. The Legal Outcome: A former Google software engineer found guilty of stealing the tech giant's trade secrets was sentenced to just under a year in prison Tuesday [September 1, 2026]. While a federal judge upheld the theft of trade secrets conviction, part of his economic espionage conviction was overturned due to insufficient evidence that he knew his conduct directly benefited the Chinese state government. Beyond traditional corporate espionage, U.S. officials and companies have sounded the alarm on an "industrial-scale" practice known as model distillation. How It Works: Distillation involves querying an advanced, frontier-level U.S. AI model millions of times and analyzing the outputs. Foreign labs then use that data to train their own cheaper, smaller, and highly capable copycat models without paying for the immense R&D costs. Accused Entities: Major American AI developers, including Anthropic and OpenAI, have explicitly accused Chinese firms of "free-riding" on their models. Anthropic singled out Chinese AI labs like Moonshot AI (specifically its Kimi K3 model), DeepSeek, MiniMax, and tech giant Alibaba for orchestrating massive distillation attacks to steal the capabilities of their Claude models The scale of the threat has forced unprecedented collaboration within the competitive tech sector and prompted aggressive federal action.
Have a great month!
Fernando