DeepSeek is developing an inference-focused AI chip to cut reliance on Nvidia and Huawei, Reuters reports. The move follows a large funding round and U.S. scrutiny; success would deepen a cost advantage, but technical and geopolitical risks remain.
A TikTok video quickly turned OpenAI’s new GPT‑Live‑1 voice model into a viral embarrassment after the system failed a trivial letter‑count test, raising fresh questions about voice AI reliability.
Will Samsung’s 18‑fold profit surge mark a durable re‑rating? AI demand for DRAM and HBM is lifting near‑term results, but memory’s volatile cycles and rivals’ gains complicate the outlook.
Chinese firms are racing to build human‑like robotic hands. Startups claim multi‑joint designs and seed funding, but experts warn sensing, durability and real-world success remain elusive.
A deal promoted on Mashable and PCMag offers lifetime access to Claude, Gemini and GPT via ChatPlayground for $55.30 with code JULY30, but the offer relies on third‑party arrangements and lacks independent verification.
Boston Dynamics’ humanoid Atlas handed the match ball at Brazil v. Norway on July 5, 2026, a Hyundai-backed World Cup activation. The stunt shows polished hardware but not full autonomy.
Macron and Modi have personally lobbied tech CEOs to anchor AI data‑centre and cloud investments in France and India. SoftBank, Amazon and Google made headline pledges; the crucial question is whether permits, grids and local buy‑in will convert promises into capacity.
As humanoid robots move into warehouses and offices, makers like Agility Robotics are marketing safety features alongside dexterity. But engineers, buyers and insurers still lack standardized tests to prove those promises.
FIDE suspends former world champion Vladimir Kramnik after he publicly accused players, including Daniel Naroditsky, of cheating. Reporting conflicts on the exact ban length and FIDE has not published the full disciplinary ruling.
Yahoo Finance says recent 10–20% pullbacks create a 'second-chance' to buy AI stocks but omits the two recommended tickers. Analysts warn valuations and IPO activity complicate the trade.
Could recent grads be AI’s first workforce test? Goldman Sachs data cited by Yahoo Finance suggests degree holders face earlier exposure because they cluster in finance and professional services. A Reuters/Ipsos poll shows 50% of grads use AI regularly.
WIRED reports Meta ran a contractor program called "Cannes" that paid people to pose as teens and feed rival chatbots thousands of sexual, suicidal and disturbing prompts. Meta calls it safety testing; critics call the scale and method alarming.
