⚡Big Tech’s Carbon Retreat Could Become Climate’s Next AI Casualty
Plus: Huawei, DeepSeek, and the Big Threat.
As tech companies struggle to keep up with rising energy demand, they’re doing whatever it takes to protect margins, meet investor expectations, and maintain a competitive edge. That now includes scaling back carbon emission goals and purchasing fewer carbon credits, raising a bigger question: where does this lead long term? At the same time, Huawei is emerging as a serious challenger to Nvidia and broader U.S. chipmakers, signaling intensifying global competition in AI infrastructure. Let’s dive in and stay curious.
Big Tech’s Carbon Retreat Could Become Climate’s Next AI Casualty
🧰 AI Tools - Autonomous Agents
Huawei, DeepSeek, and the Big Threat.
📚Learning Corner - The Complete Perplexity Mastery Guide
📰 AI News and Trends
Anthropic shifts its enterprise pricing to charge businesses based on the amount of AI they actually use. This will add the client’s cost, especially those using Claude code.
Claude now allows you to create polished visual work like designs, prototypes, slides, one-pagers, and more, a feature to compete with Figma.
Google has assembled a strike team of researchers and engineers to improve its AI coding models, automate more of its own coding, and ultimately improve its AI research.
Cursor is in talks to raise over $2 billion at a $50 billion valuation, led by investors like Andreessen Horowitz and Thrive Capital.
Other Tech News
QVC Group, owner of QVC and HSN, will file for Chapter 11 bankruptcy protections. The internet has killed another business. Audiences that once watched sellers on TV have moved to 24/7 livestream platforms like Whatnot, TikTok Shop, or Temu.
Meta targets May 20 for the first wave of layoffs, which will be ~8000 layoffs; additional cuts later in 2026
An upcoming AI-generated movie about Bitcoin, of course, a development technology, is still developing a protocol, which knocked production costs down from $300 million to around $70 million
Sam Altman may get fired, again… OpenAI Investors Aren’t Sure Sam Altman Is the Guy to Take Them Public
Uber will now return your packages; no need to leave your home or office.
Gucci-branded Google smart glasses are coming next year
Big Tech’s Carbon Retreat Could Become Climate’s Next AI Casualty
Microsoft’s reported pullback from future carbon removal purchases is the clearest sign yet that the AI boom is colliding with Big Tech’s climate promises. According to Heatmap, Microsoft has told some suppliers and partners it would pause future purchases, although the company says its carbon removal program has not ended and that it may simply adjust pace and volume as it refines its strategy. Even that softer wording matters. Microsoft has been the backbone of the young carbon removal market, accounting for roughly 87% of purchases in 2025 and 56% in the first quarter of 2026, while Heatmap reported it made about 90% of global purchases last year.
The pressure is coming from electricity. As companies race to build AI data centers, their power demand is surging faster than clean energy supply can keep up. The Associated Press reported that Google’s emissions have risen nearly 50%, Amazon’s 33%, Microsoft’s more than 23%, and Meta’s more than 60% as the sector struggles to reconcile years-old climate pledges with the immediate need for more computing power. In that sense, Microsoft is not alone. It is the most visible company stepping back from carbon removal buying, but Google, Amazon, and Meta are also showing how AI expansion is weakening climate commitments in practice, even if they have not formally abandoned them.
This is a big issue for two reasons. First, carbon removal was never supposed to replace emissions cuts, but climate scientists say it is still necessary to reach global temperature goals. If the largest corporate buyer slows down, a fragile industry loses the demand signal it needs to scale. Second, if tech companies decide that power growth matters more than climate follow-through, they risk locking in more gas-fired electricity, higher emissions, and a market message that sustainability promises can be postponed when AI profits are at stake. That would hurt not only carbon removal startups, but also the credibility of corporate climate pledges across the economy.
🧰 AI Tools of The Day
Autonomous Agents
AgentGPT - A browser-based autonomous agent tool that lets users launch agent-like workflows that plan and execute tasks from a prompt.
BabyAGI - A lightweight autonomous task management agent that allows “task creation, prioritization, and execution” style that made agent tools popular.
OpenDevin - An open-source autonomous software agent project that aims to let an AI agent perform real actions in a working environment, especially for technical tasks.
SuperAGI -An open-source AI agent framework and platform that focuses on building, running, and managing autonomous agents with tools and memory.
Huawei, DeepSeek, and the Big Threat.
In a recent interview, Jensen Huang pointed out that Huawei is becoming a threat to Nvidia and to the US tech landscape entirely. Huawei has not made a better chip than Nvidia. And it may not be near that. The more important point here is that Huawei may be becoming good enough, especially when paired with a leading Chinese AI company such as DeepSeek. That is what makes Jensen Huang’s warning significant. If DeepSeek can run a major model on Huawei’s Ascend platform, then China no longer needs to match Nvidia chip-for-chip to weaken Nvidia’s position or reduce America’s leverage over the AI stack.
What makes this a potential blow to Nvidia is that Nvidia’s moat has never been just about having the fastest silicon. Its real strength has been the combination of chips, software, developer tools, and industry lock-in. Researchers build for CUDA because CUDA is the default. Startups buy Nvidia because its software was written for Nvidia. Cloud providers stock Nvidia because customers expect Nvidia compatibility. Once a serious Chinese lab proves that frontier-class models can be adapted to Huawei’s stack, that lock-in begins to weaken.
That matters beyond DeepSeek itself. DeepSeek is important because it can act as a proof of concept for the rest of China’s AI sector. If one of the country’s most visible AI firms shows that Huawei hardware is commercially usable, then other Chinese companies may decide that the cost of switching away from Nvidia is worth paying. Even if Huawei chips remain less powerful, they can still become strategically decisive if they are sufficiently available, sufficiently optimized, and supported by a growing domestic software ecosystem. In technology markets, good enough plus local control can be more disruptive than a technically superior but politically constrained foreign supplier.
The foldable-phone example matters for the same reason. Huawei’s launch of a widely available premium foldable ahead of Apple and Samsung is not directly an Nvidia story, but it strengthens the broader Huawei narrative: this is a company that sanctions failed to kill. Instead of being frozen out of the market, Huawei has re-emerged as a company capable of setting the pace in selected categories. That matters because perception influences procurement, partnerships, and state backing. A company that looks like a survivor in smartphones is more likely to be trusted as a long-term platform in AI infrastructure as well.
For U.S. companies more broadly, the danger is not only lost sales. The danger is the emergence of a parallel technology system. If Huawei supplies chips, operating systems, cloud infrastructure, and flagship devices, then Chinese firms may no longer need American suppliers in the most strategically important layers of the stack. NVIDIA loses accelerator demand. U.S. cloud companies lose influence. American software ecosystems lose default status. And Washington loses part of the coercive power that came from controlling the tools on which advanced AI depended.
This is why Huang’s comment is so striking. He is effectively saying that the real risk is not that Huawei becomes better than Nvidia next quarter. The risk is that Huawei and DeepSeek together prove that China can build an AI path that is independent enough to be viable. Once that happens, export controls stop being a choke point and start becoming an incentive for China to accelerate replacement. In that scenario, Huawei’s recent wins are not isolated product victories. They are signs that an alternative AI and consumer-tech ecosystem is becoming more real.