Technology Tales

Advanced problem-solving models, known as reasoning models, have been developed to perform complex tasks such as coding, scientific reasoning and multistep planning. These models think before responding, producing a chain of internal thought before generating an answer, and are particularly useful for tasks that benefit from high-level guidance rather than precise instructions.

The models use reasoning tokens, which are not visible to the user, to break down prompts and consider multiple approaches before generating a response. To manage costs, it is possible to limit the total number of tokens generated, covering both reasoning and completion tokens. Ensuring sufficient space in the context window for reasoning tokens is important, as failing to do so can result in costs being incurred without a visible response being returned.

The models are accessible through various endpoints, and developers may need to complete organisation verification before accessing certain options. When prompting these models, providing high-level guidance and allowing them to work out the details themselves tends to produce better results than supplying overly precise instructions.

To securely and reliably allow traffic from ChatGPT agents to reach a site, authentic traffic can be identified by checking for specific headers. The ChatGPT agent signs every outbound HTTP request, enabling confident identification of genuine traffic.

This is achieved through the use of HTTP Message Signatures, which include a Signature and Signature-Input set of headers alongside a companion Signature-Agent header. By verifying these headers and checking the public key associated with the signature, it is possible to confirm the authenticity of the request.

Cloudflare users can allowlist ChatGPT agent traffic by creating a rule that skips or allows requests from verified bots. Users of other CDNs can achieve the same result by checking the request headers and verifying the signature directly.

ChatGPT Agent is a feature that enables ChatGPT to complete complex online tasks on behalf of users, including conducting research, filling out forms and editing documents, whilst allowing users to remain in control. Access requires a subscription to certain plans, such as Pro, Plus or Team, and the feature is available across web, mobile and desktop apps. It is not currently available in Switzerland or the European Economic Area, though expanded access is expected.

Users can schedule tasks to repeat and can view and manage their tasks through the interface. The feature includes safeguards to help prevent privacy risks such as prompt injection attacks.

To keep data safe, users are advised to exercise caution when logging in to websites or using connectors. Recommended best practices include not typing passwords or private information directly into messages and regularly reviewing connector permissions. The feature takes screenshots to interact with web pages but does not capture sensitive data when users are controlling the virtual browser. User data are handled in accordance with the provider's privacy policy, and chats and screenshots are retained until deleted by the user.

DeepLearning.AI

DeepLearning.AI is an online education platform founded by Andrew Ng in 2017 with the aim of making top-tier artificial intelligence education accessible globally. The platform offers a wide range of courses and certifications covering deep learning foundations, natural language processing and AI for non-technical audiences.

Led by Andrew Ng, a prominent figure in artificial intelligence and a consistent advocate for accessible AI education, the platform combines expert instruction with hands-on projects and a supportive community. Its stated mission is to democratise AI tools and skills for broad societal benefit.

Sudowrite is an AI-powered writing tool designed specifically for fiction authors, offering a suite of features to assist at every stage of the creative process. It can generate suggested prose in the writer's own voice, expand underdeveloped scenes, rewrite passages, and provide actionable editorial feedback across multiple drafts. Additional tools allow writers to brainstorm names, plot points and ideas, explore character arcs and themes on a visual canvas, and generate artwork from descriptive writing. The platform supports full novel development from initial concept through to completed chapters, and extends its functionality through a plugin system offering over a thousand additional capabilities.

It has received favourable coverage from publications including The New York Times and has drawn endorsements from bestselling authors and award-winning screenwriters. The company was founded by writers Amit Gupta and James Yu, with backing from figures connected to platforms such as Medium and WordPress, as well as a number of notable film writers and directors.

Anthropic has unveiled a new Integrations feature enabling Claude to connect with various applications and tools, alongside an enhanced Research capability that can search the web, Google Workspace and integrated apps. The Research function allows Claude to investigate topics for up to 45 minutes before delivering comprehensive reports with proper citations.

Initially available to users on premium plans, Integrations supports ten popular services including Atlassian's Jira, Zapier, Cloudflare and Intercom, with more partnerships forthcoming. Developers can create their own integrations in approximately 30 minutes using the provided documentation.

These updates expand Claude's functionality by allowing it to understand project histories and organisational knowledge, and to take actions across multiple platforms. The result is a more capable and context-aware tool for complex project management tasks.

The integration of data science and artificial intelligence is transforming biometrics careers, with employers now valuing candidates who possess hybrid skills, are familiar with newer platforms and can adapt to complex data environments. As organisations adopt more automated systems and predictive modelling tools, traditional biometric roles are being redefined, with a greater emphasis on interpretation, validation and system-level oversight.

Biometrics teams must be able to work alongside automated systems, validate outputs and ensure that data meets regulatory standards. Programming fluency, experience with cloud tools and familiarity with machine learning libraries are becoming increasingly important in this context.

Employers should prioritise candidates who understand regulated systems and can support traceability and inspection readiness. They should also provide training on audit trail review, output validation and the documentation of overrides in order to upskill their existing teams. The most effective biometrics teams will combine strong analytical skills with a clear understanding of how automated outputs must be validated, interpreted and documented to meet regulatory standards.

Large language models are inherently non-deterministic, meaning they can produce different responses to the same input, which can lead to errors and inconsistencies. This lack of determinism can be problematic in enterprise software applications where reliability is crucial. To mitigate this issue, developers can implement measures such as sanitising inputs and outputs, observing the process as much as possible and ensuring that processes run once and only once. Additionally, using durable execution technologies can help save progress in workflows and prevent repeated calls to external services. By introducing these controls, developers can make large language models more reliable and trustworthy, which is essential for building robust enterprise software applications that organisations can rely on.

SingleStore is a distributed database platform built for low-latency SQL, designed to handle high user concurrency and petabyte-scale datasets whilst delivering fast query performance across analytics, transactions and search within a single unified engine. The platform includes an AI capability called Aura, which encompasses an AI-powered analyst tool that converts natural language into SQL queries, alongside functions for machine learning and large language model integration. These enable tasks such as sentiment analysis, summarisation and data classification directly within SQL.

The platform is used by a broad range of major organisations including Adobe, Cisco, Goldman Sachs and NVIDIA, with users citing its performance, ease of use and cost-effectiveness as key strengths. SingleStore integrates with existing data infrastructure and is positioned as a solution for industries with demanding real-time data requirements, including financial services, media and high-tech sectors.

Teradata is a technology company offering an AI and data platform designed to help large enterprises convert organisational knowledge into autonomous, real-time business decisions. Its platform combines agentic AI, connected data and scalable analytics to serve industries including financial services, healthcare, manufacturing, retail and telecommunications.

The company counts among its clients ten of the top ten global banks, seven of the top ten airlines and seven of the top ten telecommunications providers. Notable examples of its work include helping Mitsui Sumitomo Insurance increase contract rates by 250% and supporting OSF HealthCare in achieving a 520% increase in patient engagement.

Teradata operates through a flexible, consumption-based pricing model and delivers its services via cloud, on-premises or hybrid deployments. The platform is supported by a partner ecosystem that includes major technology providers such as Microsoft, Google, AWS and NVIDIA.