OpenAI gets $110 billion in funding from a trio of tech powerhouses, led by Amazon
https://apnews.com/article/openai-amazon-nvidia-softbank-altman-microsoft-a0a915c32b85337d799fe2f9525a932a
$110 billion in funding for OpenAI sounds like a tech utopia jackpot, but this colossal bankroll is a double-edged sword. The concentration of capital from Amazon, Nvidia, and the usual suspects only deepens oligopolistic control over AI’s future, stifling diversity of innovation and locking out smaller players. Meanwhile, the sheer scale raises questions about accountability and the true intentions behind this hyper-investment—are we funding safe AI development or corporate dominance disguised as progress?
Perplexity signs a multiyear deal with CoreWeave to use dedicated clusters powered by Nvidia Grace Blackwell chips for AI inference; CRWV jumps 5%+ pre-market
https://www.axios.com/2026/03/04/perplexity-coreweave-data-center-nvidia
Perplexity’s reliance on Nvidia’s Grace Blackwell chips via CoreWeave highlights the semiconductor bottleneck strangling AI scalability. This deal fuels Nvidia’s chokehold on AI infrastructure, which is less a sign of a competitive ecosystem and more an indication of tech monoculture risk. If Nvidia stumbles or decides to leverage pricing power aggressively, AI companies dependent on these specialized chips face systemic vulnerabilities that investors are quietly ignoring.
AMD Ryzen AI 400 chips will bring newer CPUs, GPUs, and NPUs to AM5 desktops
https://arstechnica.com/gadgets/2026/03/amd-ryzen-ai-400-cpus-will-bring-upgraded-graphics-to-socket-am5-desktops/
AMD’s Ryzen AI 400 promises shiny new hardware but masks a stagnating innovation cycle in desktop AI chips. This incremental upgrade plays into the hands of entrenched x86 architecture dominance, ignoring the rise of domain-specific accelerators and custom silicon disrupting traditional CPU-GPU paradigms. The desktop market’s obsession with minor generational improvements risks delaying the shift toward more efficient, specialized AI compute frameworks that can break current energy and performance walls.
Dario Amodei says Anthropic has a better retention rate than OpenAI, reminding its employees of its “mission” while countering competitors trying to hire them
https://www.theinformation.com/briefings/anthropic-ceo-dario-amodei-says-mission-helps-fend-rivals
Anthropic’s touted “mission” retention strategy is less about idealism and more a subtle psychological play to mask the hyper-competitive, talent-scarce AI labor market. The narrative of mission-driven loyalty obscures the brutal counteroffers and poaching wars beneath the surface, indicating that cultural branding is now weaponized HR. This battlefield for AI talent risks creating echo chambers rather than fostering the diverse, cross-pollinated innovation necessary for genuine breakthroughs.
Self-driving software startup Oxa raised a $103M Series D, with $50M coming from the UK government’s National Wealth Fund; Nvidia’s NVentures also invested
https://sifted.eu/articles/oxa-raises-103m-nvidia-bp
Oxa’s funding round reveals an alarming fusion of government and corporate interests in AI-driven autonomy, blurring lines between public policy and private profit. The UK’s National Wealth Fund investment signals a geopolitical race for control over self-driving tech, but also exposes taxpayers to the volatility of venture capital and the uncertain regulatory landscape. Nvidia’s involvement further entrenches its grip on the AV tech stack, raising red flags about dependency and potential monopolistic behavior in a sector critical to future mobility.
Sources: Meta is creating a new applied AI engineering organization that will have an ultra-flat structure and help bolster Meta’s superintelligence efforts
https://www.wsj.com/tech/ai/meta-to-create-new-applied-ai-engineering-organization-in-reality-labs-division-d41c4a69
Meta’s “ultra-flat” AI engineering org is corporate speak for rapid decision-making without the usual checks and balances, a governance shortcut that could accelerate breakthroughs—or catastrophic missteps. The push for “superintelligence” inside Reality Labs signals a reckless escalation in AI ambition, but the lack of transparency and oversight raises urgent questions about ethical guardrails. This structure risks prioritizing speed over safety in a domain where the stakes are existential.
Sources: Hacker News, Techmeme, AP News, Ars Technica | Compiled March 04, 2026