ByteDance Rolls Out Cici Chatbot in UK, Mexico and Southeast Asia via Ads and Influencers
ByteDance promotes Cici abroad, using quirky ads and creators to boost its cartoon chatbot’s fame, but questions remain for investors…
ByteDance’s Doubao Surpasses DeepSeek, Wins Tens of Millions by Putting Usability First
ByteDance’s Doubao reclaimed attention from DeepSeek, proving friendlier design beats raw model claims, and a surprise looms ahead for rivals…
MIT Study: Giant AI Models May Soon Hit Diminishing Returns, Smaller Efficient Models Gain Traction
Massive AI model bets may falter if efficiency improvements shift advantage; cloud giants might rethink spending, and then what happens…
Ant Group debuts Ling-1T, a trillion-parameter open-sourced model that scores 70.42% on AIME and matches top reasoning efficiency
Ant Group’s Ling-1T shocks AI watchers with trillion-parameter scale and AIME prowess, promising surprises and a reveal that changes forever…
Oracle and NVIDIA launch OCI Zettascale10, delivering 16 zettaflops to power enterprise AI
Oracle and NVIDIA’s Zettascale10 cluster promises mind-bending AI scale and speed yet a single reveal could rewrite industry rules soon…
Salesforce pledges $15 billion to San Francisco to jumpstart enterprise AI, fund incubator and create jobs
Salesforce pledges $15 billion to San Francisco for AI labs, jobs, and local startups—can the city keep pace with competitors…
AI spending set to climb 32% by 2026 as 89% of businesses report no customer value
UK firms plan 32% more AI spending, yet 89% report no customer value—why are they betting big when results remain…
Meta Pushes Metaverse Teams to Integrate AI Into Every Workflow to Go 5X Faster
Meta’s metaverse chief orders AI-driven work to quintuple productivity; engineers scramble, deadlines shift, and then a secret roadmap leaks overnight…
Cisco Debuts 51.2 Tbps Router to Tackle AI Data Center Interconnect Bottleneck
Cisco’s 8223 offers 51.2 Tbps via Silicon One P200 to link AI across data centers, yet test results reveal surprises…
Samsung’s 7-Million-Parameter Model Topples Massive LLMs on ARC-AGI Reasoning Benchmark
A tiny 7-million-parameter model out-smarts giant LLMs on tough reasoning tests, and when researchers probed deeper they found puzzling anomalies…
