Sales of the iPhone hit an all-time high in the final three months of last year, tech firm Apple reported on Thursday. Revenue rose by 16% compared to the same period last year to $144bn (£82.5bn) - the strongest growth since 2021 - thanks to a jump in sales in China, as well as Europe, the Americas, and Japan. However, sales in other parts of the company were less positive. Wearables and accessories, which include things like the Apple Watch and AirPods, fell by roughly 3%. Apple chief executive Tim Cook said the iPhone's boost in sales meant the firm was in supply chase mode.

We present MedARC's team solution to the Algonauts 2025 challenge. Our pipeline leveraged rich multimodal representations from various state-of-the-art pretrained models across video (V-JEPA2), speech (Whisper), text (Llama 3.2), vision-text (InternVL3), and vision-text-audio (Qwen2.5-Omni). These features extracted from the models were linearly projected to a latent space, temporally aligned to the fMRI time series, and finally mapped to cortical parcels through a lightweight encoder comprising a shared group head plus subject-specific residual heads. We trained hundreds of model variants across hyperparameter settings, validated them on held-out movies and assembled ensembles targeted to each parcel in each subject. Our final submission achieved a mean Pearson's correlation of 0.2085 on the test split of withheld out-of-distribution movies, placing our team in fourth place for the competition. We further discuss a last-minute optimization that would have raised us to second place. Our results highlight how combining features from models trained in different modalities, using a simple architecture consisting of shared-subject and single-subject components, and conducting comprehensive model selection and ensembling improves generalization of encoding models to novel movie stimuli. All code is available on GitHub.

Chipmaker Nvidia reported its latest financial results on Wednesday, recording 30.04bn in revenue over the past three months – a 122% jump from the year prior – and showing that artificial intelligence investment mania shows no signs of cooling. Analysts had anticipated about 28.7bn in revenue. Shares slid more than 3% in after-hours trading. "The company continues to benefit from a market paradox: big tech's aggressive AI investment strategies drive massive demand for Nvidia's chips, even as these same companies invest in developing their own silicon," said Jacob Bourne, a technology analyst with Emarketer. Nvidia has told customers that its next-generation AI chips, code-named Blackwell, will be delayed several months from January, though early samples are shipping to a small group of customers now.

Nvidia reported record quarterly revenue Wednesday on the back of the explosion in corporate appetite for artificial intelligence. "The next industrial revolution has begun – companies and countries are partnering with Nvidia … to produce a new commodity: artificial intelligence," said Jensen Huang, founder and CEO of Nvidia. The company brought in 26bn in revenue in the first quarter of fiscal year 2025, up 18% from Q4 and up 262% from a year ago. Net profit was 14.88bn, up from 2bn a year before. The AI chip maker, whose fortunes are interpreted as a bellwether for the AI transformation under way, reported earnings per share were 5.98, up 21% from the previous quarter and up 629% from a year ago.

Graph anomaly detection (GAD) has gained increasing attention in recent years due to its critical application in a wide range of domains, such as social networks, financial risk management, and traffic analysis. Existing GAD methods can be categorized into node and edge anomaly detection models based on the type of graph objects being detected. However, these methods typically treat node and edge anomalies as separate tasks, overlooking their associations and frequent co-occurrences in real-world graphs. As a result, they fail to leverage the complementary information provided by node and edge anomalies for mutual detection. Additionally, state-of-the-art GAD methods, such as CoLA and SL-GAD, heavily rely on negative pair sampling in contrastive learning, which incurs high computational costs, hindering their scalability to large graphs. To address these limitations, we propose a novel unified graph anomaly detection framework based on bootstrapped self-supervised learning (named BOURNE). We extract a subgraph (graph view) centered on each target node as node context and transform it into a dual hypergraph (hypergraph view) as edge context. These views are encoded using graph and hypergraph neural networks to capture the representations of nodes, edges, and their associated contexts. By swapping the context embeddings between nodes and edges and measuring the agreement in the embedding space, we enable the mutual detection of node and edge anomalies. Furthermore, BOURNE can eliminate the need for negative sampling, thereby enhancing its efficiency in handling large graphs. Extensive experiments conducted on six benchmark datasets demonstrate the superior effectiveness and efficiency of BOURNE in detecting both node and edge anomalies.

Brands are scrambling to incorporate generative AI into their strategy to stay ahead of the curve. But according to our analysts, AI's current uses are just the tip of the iceberg. Here are some recent AI innovations and predictions for the tech's evolution. Last month, Snapchat launched My AI, a chatbot powered by OpenAI's ChatGPT technology. The chatbot can recommend birthday gift ideas, plan a hiking trip, or suggest a dinner recipe.

One of the challenges of aerial imagery, whether from an airplane or a satellite, is making sense of what you see. What is that image telling you? Ceres Imaging, a California startup with offices in Nebraska and Washington, is using artificial intelligence to answer that question. The company is entering its ninth crop season of providing high-resolution crop imagery for customers. However, John Bourne, vice president of marketing, Ceres Imaging, says the company wanted to work on ways to "productize" the good science it was developing, so three years ago it brought artificial intelligence technology to irrigation issue identification.

CUYAMA, CA - APRIL 28: Overhead irrigation of this newly planted crop of carrots is putting ... [ ] pressure on the available groundwater supplies as viewed on April 28, 2020, in Cuyama, California. Located in the northeastern corner of Santa Barbara County, the sparsely populated and extremely arid Cuyama Valley has become an important agricultural region, producing such diverse crops as carrots, pistachios, lettuce, and wine grapes. The global precision farming market includes technology like robotics, imagery, sensors, artificial intelligence (AI), big data and bio-engineering is expected to reach more than $16 billion by 2028, according to a March 2021 report from Grand View Research. What if you could combine AI and traditional aerial imagery to build data sets that help farmers and food processors gain insight into crop heartiness while it was still growing in the field? Saul Alarcon, an Agronomist at The Morningstar Company that sources and processes tomatoes for several tomato-based products, says that new agriculture technologies based on AI can improve farming decisions.

Computer historians have staged a re-enactment of World War Two code-cracking at Bletchley Park. A replica code-breaking computer called a Bombe was used to decipher a message scrambled by an Enigma machine. Ruth Bourne, a former wartime code-cracker who worked at Bletchley and used the original Bombes, oversaw the modern effort. Enigma machines were used extensively by the German army and navy during World War Two. This prompted a massive effort by the Allies to crack the complex method they employed to scramble messages.

Static knowledge about gait and anatomy is represented in frames and dynamic evaluation strategies are represented in frames and metarules. Initial results are described by Dzierzanowski et al. (Dzierzanowski et al., 1983). We have completed several expert systems for electroencephalogram evaluation (Jagannathan, et al., 1981, 1982) (Bourne et al., A rule-based consultant system has been implemented for advising physicians about the prescription of initial dialysis therapies (Schaffer et al, 1983). This system is now in use at the Dialysis Clinics, Inc., Nashville, Tennessee. This system is now being expanded into a community of simulated consultative experts that provide advice about pharmacology, cardiovascular problems, nutrition and other problems Personnel: J. D. Schaffer, J. Cavaedes, J Bourne This project is devoted to building a complete system that assists the electromyogram [EMG] reader.