Stripe on Monday released a preview of a new feature that could help AI startups (and other companies) solve the problem of passing through the underlying costs of AI model usage to their customers.
Stripe’s feature, however, goes even further than just passing through the costs of the tokens. It allows startups to charge a markup percentage on token usage. So a company can, for instance, charge an automatic 30% above the cost of the tokens that the startup will pay the model maker.
As Stripe described it, “Say you’re building an AI app: you want a consistent 30% margin over raw LLM token costs across providers. Billing automates the process.”
The billing feature lets the startup pick the AI models it uses. It tracks the API prices of those models. It then records the customers’ token usage and applies the profit-margin markup automatically.
As we’ve previously reported, there are a variety of ways that AI startups are charging for their wares. Many of them charge tiered monthly subscriptions that have usage-rate caps; once those are hit, the subscriber may be charged more for exceeding the limit.
For instance, Cursor last year changed the pricing on some of its tiers from unlimited use to rate-limited usage, with fees for extra consumption on top.
Without a usage cap, users could run up big bills for a startup with the model makers, and force the startup to operate in the red. This is especially acute for agentic startups. The more their customers use their agents, the more tokens they consume from the underlying model provider, be that OpenAI, Google Gemini, Anthropic or others — making pricing and business model decisions especially critical.
Techcrunch event
San Francisco, CA | October 13-15, 2026
Stripe has also introduced its own AI gateway, a tool that give users access to multiple models, letting them choose the best one for the job. But the billing tool also works with third-party gateways that are already popular, like those offered by Vercel and OpenRouter, according to a tweet by a Stripe product manager,
There are, of course, other startups offering AI model cost management features with their own gateways. OpenRouter, for instance, which grants access to over 300 models, charges a flat 5.5% markup over the token fees for its first-tier plan, and offers budget controls, too.
Stripe is not currently charging its own markup on the gateway, its product manager said on Twitter. The feature, however, is still in waitlist mode. Either way, if Stripe can help startups easily turn tracking and billing for this expense into a profit-maker, it could be a game-changer. Stripe did not immediately respond to a request for comment on when the feature may be generally available.
The AEC industry is no stranger to outsourcing. Back in the day, architectural design firms probably saw outsourcing as a quick alternative to full-time hiring in times when the workload was too much to handle by the in-house team. They may hire smaller firms or freelancers to do the drafting or basic quantity takeoffs on smaller projects, just to keep the workflow running while the internal team is preoccupied with higher-stakes tasks. Things have changed quite a bit since then, most likely as soon as 3D modeling turned mainstream and BIM entered the scene afterward.
With the possible exception of real-time rendering in architectural visualization, BIM is arguably the biggest thing to have ever happened to the industry in a long time. It has introduced a new paradigm in design workflow by putting everything into a data-rich digital simulation where structural designs, materials, scheduling, cost estimation, clash detection, and energy performance are laid bare. The interactive nature of the BIM file makes the architectural design process an immersive experience and is loaded with all the data you need to formulate an informed design decision each time. The problem is that BIM professionals are few and far between, at least for now.
Outsourcing, which started as an “option” in the old days, has now become a necessity if an architectural firm wishes to take full advantage of everything BIM has to offer. And as an AEC-focused freelancing platform, Cad Crowd has the resources and the network to connect you with some of the most talented BIM specialists from all over the world.
🚀 Table of contents
BIM outsourcing strategies
With the lack of BIM experts in the industry, outsourcing is no longer a simple cost-saving measure for architectural firms, but a strategic move to gain a true competitive advantage. There are more than a few ways to make sure you get the best out of your money by hiring external BIM experts; some are listed below.
Pilot project
As a general rule, it’s a bad idea to fall into the trap of hiring a BIM partner solely based on what you see on their portfolio and other clients’ reviews. While both are pretty good indicators of quality of services, you can certainly use a little bit more concrete evidence of their capabilities. BIM is as sophisticated as architectural drafting services get at this point, and most likely used for a reasonably large project that involves complex technical calculations of geometry, bill of materials, prefabricated components, etc. And just because a potential partner, whether a firm or an individual freelancer, says they can do the job well, you are under no obligation to take them at their word. Instead, use the “pilot project” strategy.
The idea behind a pilot project is to gauge the partner’s technical proficiency and see if they have the expertise to handle a small-scale BIM task. In addition, you also get the chance to establish an efficient communication protocol without any big investment. Since the primary objective is to make sure that the partner can work at the pace you desire, hire them to do a BIM project that’s somehow representative of your typical workflow. The project must be small enough that you don’t have to spend too much money on it, yet sufficiently challenging because it needs to be a test at the same time.
Another benefit is that you have the chance to establish a BIM Execution Plan (BEP) early on, in case you actually end up collaborating with the partner for future and larger projects. BEP may contain guides to procedures like file naming conventions, software to use, file formats, data sharing, and so forth (you know a project is complex enough if you have to use an elaborate execution plan). Assuming the pilot project is successful with the partner delivering exactly what’s promised, you gain the confidence to bring the collaboration further. If not, move on to the next candidate. Either way, the time and effort you spend running the pilot project are resources well spent.
Outsourcing makes little sense if the deliverables fail to meet your expectations. In an ideal world, you should only outsource a BIM project to a partner proficient enough to create models that exceed (or at least match) the in-house quality standards. Otherwise, you might as well just rely on the team you already have.
It’s only natural that most freelancers and BIM service providers claim to have their own internal quality assurance professionals whose primary role is to ensure that all files are checked for errors and analyzed for inconsistencies before delivery to clients. All this sounds very reassuring, up until the point you remember that true professionals run their work through a third-party evaluation service. There’s no need to hire any independent analyst here; if the partner does use such a service, it’s a big plus all to your advantage, but you still need to practice due diligence anyway. An effective analysis happens in a layered review process to make sure that the audit is dependably objective:
1st Layer: an automated review process in the form of clash detection by software. Solibri, Navisworks Manage, Revizto, ClashMEP, Autodesk Construction Cloud, BIM Track, and Bentley Navigator are among the most widely used options.
2nd Layer: Make it clear in the project brief that the deliverable must be reviewed by at least one senior member of the outsourced team. Provide a list of the quality standards used by your firm and have the partner compare it against the deliverables.
3rd Layer: your in-house team reviews the work, both automatically and manually, to ensure everything is in order. This means the outsourced team has to send a copy of the file ahead of schedule to compensate for the audit process and the possibility of a revision or two.
Please note that revisions are nothing but normal in any kind of outsourced task. The lack of direct supervision and management means you can’t control everything that happens during the workflow. Just because the deliverable isn’t 100% correct doesn’t mean the outsourced team is doing a bad job. In some cases, multiple rounds of revisions are still acceptable so long as the mistakes are within the margin of error.
BIM component outsourcing
Considering how popular BIM is in today’s architectural landscape, there’s a good chance that your architectural planning and design firm will gradually (if not already) implement the technology in current and future projects. In the grand scheme of things, however, adoption is slow. Even for firms that have used BIM for a little while at this point, they still occasionally suffer from what’s typically referred to as BIM Bloat.
One thing that separates BIM files from other architectural visualizations is the amount of information contained directly in the models. BIM is supposed to be data-rich, filled with just about every single detail you can muster about structural elements, materials, plumbing configuration, HVAC installations, construction schedules, cost estimation, timeline projections, and more. With that in mind, there’s actually such a thing as too much data. Information overload happens when a BIM file is embedded with an excessive amount of non-critical data, leading to performance issues and unreliable error identification.
In a specialized component outsourcing scenario, you’re not hiring a partner to build an entire 3D building model from scratch. The main focus of the collaboration is to create efficient content libraries (such as Revit Families) to ensure standardized high-performance BIM components. For example, you can ask the partner to parametric components like cabinetry, doors, appliances, roofing, flooring, piping, or basically any architectural element with only the necessary metadata, such as manufacturer links, fire ratings, U-Values, etc. This project serves two major purposes: achieving a leaner component library and preparing an in-house team to create better BIM files.
Specialized BIM partner
It’s not uncommon for an architect or a small architectural firm to outsource or subcontract a certain portion of a project. A licensed architect is a qualified expert in building design, code compliance, project management, and cost estimation, but they might not be as adept at trade-specific tasks like HVAC duct routing, electrical wiring, or plumbing pipe installation. A typical strategy to solve the problem is to outsource the jobs to a BIM partner that specializes in MEP (mechanical, electrical, plumbing) and MEP drafting services.
This allows the architect to maintain control over design intent while ensuring that the actual construction/fabrication of MEP components has no negative impact on the building’s aesthetic and structural integrity. Because the MEP services will be integral parts of the building, the outsourced partner will need access to the architectural draft so they can configure a proper integration between the services and the structure itself. An MEP specialist often works directly with fabricators to locate the clearance zones (based on the provided draft) for installation and determine the right tolerances for maintenance. Once the plan is ready, the partner integrates the diagram with the architect’s BIM file for clash detection.
A big architectural firm probably employs an MEP professional to handle the task, but an architect running a small firm or a one-person business probably cannot justify hiring a full-time salaried expert for the job. Outsourcing to a specialized BIM partner enables the architect to focus on the design and management side of the project rather than getting bogged down with the particulars of non-structural building components.
Not every architectural project needs a Building Energy Modeling (or BEM), which is a subset of BIM that specifically concerns indoor air quality, energy consumption, and acoustics. But thanks to growing awareness of the importance of eco-friendly architecture and sustainability, every architectural firm must take these issues more seriously.
In short, BEM uses a digital model of a structure to evaluate energy performance under various conditions. Take, for example, an architectural model to simulate and analyze the energy-saving potential of different HVAC configurations; even seemingly simple things like insulation materials can affect the overall effectiveness of the entire service installation over a long period of use. Hiring a BEM freelancer should count as specialized CAD outsourcing, too.
Instead of hiring a sustainability consultant full-time or purchasing an expensive simulation tool that you use probably no more than half a dozen times a year, outsourcing the task to a specialized professional is the more sensible choice. Because you have to send the BIM file of the structural design to ensure accurate evaluation, such a project may involve an NDA, so make sure to prepare the form and that the outsourced partner is willing to agree to your IP protection term.
BEM should provide not just a report of the energy performance, but also simulation-based suggestions on various elements like the size of the HVAC system, ventilation, window-to-wall ratio, design of shading device, roof insulation, and more. BEM outsourcing is a popular strategy often used by small to mid-sized architectural firms to compete with their bigger counterparts. In the event a client asks about why a certain design approach is taken while the alternative is discarded, the firm can easily explain that everything is based on a valid simulation technology by a professional partner. This is how the small firms increase their credibility without allocating big financial resources – resources they can otherwise spend on the core design and construction services.
Hiring freelancers as external non-payrolled professionals makes things practical and straightforward. You pay someone from outside the company to handle tasks that would be too time-consuming or resource-demanding for the in-house team to do. Think of it as bringing in a reinforcement without all the overheads typically associated with full-time salaried employees, like benefits and insurance. In many freelance engagements, the work is performed off-site, and this is especially true for architectural drafting projects. The freelancer does much of the drafting work without direct contact/supervision from the employer. For example, a freelancer based in Canada might be hired by an architectural firm located in the United States. Sometimes, the employer and the employee are on different continents entirely.
But CAD outsourcing is also possibly done in a “staff augmentation model” method, where the employers bring in additional workforce to reinforce the in-house team. The external employees are hired on either a contractual or per-project basis, but they may work alongside the internal ones under the same roof, at least temporarily. These CAD design services serve as an extension of your firm for the duration of the contract.
A staff augmentation model makes good financial sense if the external team only consists of a small group of people, or perhaps an expert BIM consultant, and they’re locally-based. Otherwise, you may have to compensate for the travel and accommodation expenses, which can pile up to a huge bill very quickly. This is not to say that overseas staff augmentation always amounts to an expensive venture, because it all depends on how the reinforcement can improve your team and project.
Knowledge transfer
Considering how the AEC industry is in a transitory period from 2D drafting to BIM implementation, chances are you’ll see a lot of staff augmentation in the coming years in an attempt to expedite adoption. Many firms will hire BIM experts and consultants to help prepare their internal teams for the changes and to ensure a smooth transition.
One of the primary objectives of staff augmentation is to facilitate knowledge transfer. At this point, the AEC industry has to admit that trained BIM professionals are still relatively scarce. For some reason, the industry appears to be hesitant to make the move toward full adoption, despite the perceived maturity of the BIM technology itself. Many new architects on their way to licensure aren’t necessarily adept at BIM, either. In fact, many architectural schools and instructors don’t require the students to use it in the first place.
Hiring BIM experts, especially with the staff augmentation model, opens the doors to an effective hands-on knowledge transfer. BIM remains a highly specialized field in the architectural discipline. While the benefits are enormous, jumping into BIM all at once can be pretty intimidating to seasoned architects, let alone the junior ones. By including “knowledge transfer” as a job specification in staff augmentation, the freelancer understands that part of the role is to provide guidance on BIM implementation, rather than taking over the task. Guidance can be as simple as a walkthrough of a model, an hour of presentation into the BIM software at the beginning of a shift, a crash course, or anything else your firm may see fit.
The FTS workflow method
Say the in-house team is already familiar with using BIM for architectural drafting, but a project has a tight deadline that makes everyone feel rushed. A reinforcement may help reduce the workload for the team, but on the other hand, it just isn’t the right time to hire new employees. Let’s not forget that the project must be completed as quickly as possible, that even with the addition of new employees, everyone still has to work overtime almost every day just to meet the deadline.
What you need in this situation is an FTS-based outsourcing. FTS, short for “Follow-the-Sun” model, is a strategic outsourcing method that leverages the global time zones to increase productivity. The idea is to keep the project running even when the traditional work hours are already over in your local time.
Let’s say your firm is located in New York, USA, where the typical work hours last between 9 AM and 5 PM local time. Because the construction is already scheduled to begin within the next few weeks, you need to get the BIM file done in record time, as it needs to be reviewed and evaluated by the architectural design expert first. And it’s always important to take into account the likelihood of having to revise the file, too.
Outsourcing with an FTS model means hiring freelancers located in different time zones, so that the work on the BIM file is running around the clock. In this case, you need someone to pick up exactly where you left off at 5 PM. A freelancer located in Australia or Japan would make for a good candidate, because it’s still morning in those two countries while your in-house employees have already signed off. To keep the project running 24/7, you may want to consider bringing in another freelance drafter (from yet another timezone, preferably in the UK or Europe) to fill in the void between the end of the second shift and the start of the first one. The cycle continues until the BIM file is ready just in time.
Cloud-based collaboration
A prerequisite for BIM outsourcing is a cloud-based collaboration tool, and even more so if you’re using the “follow-the-sun” model. Some portions of an architectural project workflow might still rely on the tried-and-true FTP or email, but it won’t be sufficient for BIM, as it takes real-time collaboration. Among the most popular of such platforms are Autodesk BIM Collaborate Pro, Trimble Connect, BIMcollab, Graphisoft BIMcloud, Newforma Konekt, and Bentley ProjectWise.
The idea behind cloud collaboration is to allow everyone (architects, engineering design experts, designers, clients, or any authorized stakeholder) to access the BIM file simultaneously. It enables co-authoring and instant feedback because everyone is viewing, editing, and working on a single centralized dataset. Most, if not all, cloud-based platforms automatically record version history, making it easier to revert to the previous configuration in case one of the outsourced partners makes a mistake or some disproved modifications. Because changes happen in real-time, there’s minimal coordination gap.
Cloud collaboration also opens the door to effective issue tracking. Multiple teams located in different time zones might be assigned to handle specific tasks to avoid overlapping designs. If the team in the US takes care of the structural design, the freelancer in Japan can do the HVAC and MEP layout, whereas the European partner handles the issue tracking. The final design decision is ultimately at your firm’s (or the client’s) discretion, but cloud collaboration is always helpful to improve transparency. Even if you can’t provide direct supervision to overseas outsourced partners at all times, the real-time coordination keeps the workflow in control with clear deliverables, as observed in the centralized BIM file.
Every task outsourced to an external partner comes with a risk of issues concerning intellectual property and data security. And when it comes to CAD outsourcing, especially for any project that has anything to do with a BIM file, you just can’t be too careful about confidentiality. To minimize the risk, the collaboration needs to be carried out under the umbrella of compliance with ISO 19650 (the international standard for BIM information management) and Common Data Environment or CDE, to restrict file access (in practice, the outsourced partners need only to have access to the data relevant to specific tasks assigned to them).
At the very least, enforce an NDA (Non-Disclosure Agreement) as the legal framework to protect your firm against unauthorized data access/sharing of any sort.
Takeaway
There’s no denying that BIM has introduced some major improvements to the AEC industry almost in its entirety, from design and sustainability planning to construction management and cost estimation. Despite its proven usefulness, maintaining an in-house team of BIM professionals remains quite a challenge for small to mid-sized architectural firms. While software and hardware are getting more affordable every year, the relative scarcity of trained BIM specialists is a challenge too difficult to overcome at this point. It does seem that the vast majority of the available BIM workforce choose to offer their services as independent contractors instead of full-time salaried employees. And it’s no coincidence that most of those independent talents find home in Cad Crowd, a freelancing platform specializing in the Architectural, Engineering, and Construction industry. Request a quote today.
MacKenzie Brown is the founder and CEO of Cad Crowd. With over 18 years of experience in launching and scaling platforms specializing in CAD services, product design, manufacturing, hardware, and software development, MacKenzie is a recognized authority in the engineering industry. Under his leadership, Cad Crowd serves esteemed clients like NASA, JPL, the U.S. Navy, and Fortune 500 companies, empowering innovators with access to high-quality design and engineering talent.
Nothing is gearing up to launch the Phone 4a in London at a “Built Different” event set for March 5, but it gave us a sneak peek at the model’s four colorways at MWC 2026. We already knew the Phone 4a would be coming in a striking pink shade and a classic white hue thanks to prior Nothing teasers. However, this is the first time we’re seeing the official blue and black options for the Nothing Phone 4a, and it tells us a bit about the upcoming midrange smartphones.
There’s a color for everyone in the Nothing Phone 4a series, and you can check out the phone in black, blue, pink, and white colorways below. The black and white models are slightly more subtle, while the blue and pink versions are made to stand out. Seeing the Phone 4a color options in person, it’s clear that the blue is a deep, vibrant shade while the pink is a lighter, pastel hue.
Image 1 of 4
(Image credit: Brady Snyder / Android Central)
(Image credit: Brady Snyder / Android Central)
(Image credit: Brady Snyder / Android Central)
(Image credit: Brady Snyder / Android Central)
Aside from the main colors, we also saw how Nothing puts its own spin on the Phone 4a design. The phone has a transparent window that showcases a curated design underneath, with a pill-shaped camera bump in the top middle of the device.
The flash sensor is positioned off-center above the camera bump, and a ring of coils surround the entire system. This coil ring is color matched to each Phone 4a model.
We can’t forget the Glyph Bar — the latest iteration of Nothing’s Glyph Interface made for the Phone 4a series. It’s a bar with six square lights and an extra red recording light. The squares each have nine mini-LED lights for further customization. Nothing says it’s 40% lighter than the Glyph Interface in the Nothing Phone 3a series, and it’s also designed to be less distracting.
What we learned about the Nothing Phone 4a
Image 1 of 4
(Image credit: Brady Snyder / Android Central)
(Image credit: Brady Snyder / Android Central)
(Image credit: Brady Snyder / Android Central)
(Image credit: Brady Snyder / Android Central)
Looking at the Phone 4a in person, it’s clear that the six Glyph Bar squares blend in with the colorway you choose when inactive. The red recording light, however, is visible when it’s turned off. Nothing phones have typically sported a red square as part of their see-through design, so this doesn’t feel out of place. This time, it’s a functional part of the Glyph Bar system.
We also noticed that the white, blue, and pink versions of the Phone 4a have silver rear camera housing, regardless of the colorway. The black model, on the other hand, has a color-matched black rear camera housing. The black Phone 4a variant is also the only one of the group to feature black side buttons. On every other Phone 4a model, the side buttons are silver. Even on the black Phone 4a, the Essential Key is a shiny silver color to make it easily recognizable.
Get the latest news from Android Central, your trusted companion in the world of Android
For now, that’s all we know about the Nothing Phone 4a. We’ve got four fresh colorways, including two with a bit of personality. Nothing is set to officially launch the Phone 4a in a London event March 5. The new budget phones will debut alongside the upcoming Nothing Headphone a.
“At its peak in early 2014, Stack Overflow received more than 200,000 questions per month,” notes the site DevClass.com. But in December they’d just 3,862 questions were asked — a 78 percent drop from the previous year.
But Stack Overflow’s blog announced a beta of “a redesigned Stack Overflow” this week, noting that at July’s WeAreDevelopers conference they’d “committed to pushing ourselves to experiment and evolve…”
Over the past year, on the public platform, we introduced new features, including AI Assist, support for open-ended questions, enhancements to Chat, launched Coding Challenges, created an MCP server [granted limited access to AI agents and tools], expanded access to voting and comments, and more.
However, these launches are not standalone features. We have also been rethinking our look and feel, how people engage with Stack Overflow, and how content is created and shared. These new features, along with the redesign, represent how we are bringing Stack Overflow’s new vision to life and delivering value that developers cannot find elsewhere.
Our goal is to build the space for every technical conversation, centered on real human-to-human connection and powered by AI when it helps most. To support this, we are introducing a redesigned Stack Overflow to best reflect this direction… During the beta period, users can visit the beta site at beta.stackoverflow.com and share feedback as we build towards a new experience on Stack Overflow. They’ve updated their library of reusable UI components (buttons, forms, etc.), and are promising “More ways to share knowledge and ask any technical question.” (“Alongside looking for the single right answer to your question, you can now find and share experience-based insights and peer recommendations…”)
They’re launching all the planned features and functionality in April, when “More users will automatically redirect to the new site.” (Starting in April users “can continue to toggle back to the classic site for a limited time.”)
Autonomous networks — intelligent, self-managing telecommunications operations — are moving from a future vision to a current priority for telecom operators. In the latest NVIDIA State of AI in Telecommunications report, network automation emerged as the top AI use case for investment and return on investment.
Automation is different from autonomy. Beyond executing predefined workflows, autonomous networks must understand operator intent, reason over tradeoffs and decide what actions to take. Reasoning models and AI agents fine-tuned on telecom data are key to enabling this shift.
For networks to become autonomous, there’s a need for an end-to-end agentic system that includes key components like telco network models and AI agents that talk to each other and use network simulation tools to validate actions.
Ahead of Mobile World Congress Barcelona, NVIDIA unveiled an open NVIDIA Nemotron-based large telco model (LTM), a comprehensive guide for building reasoning agents for network operations, and new NVIDIA Blueprints for energy saving and network configuration with multi-agent orchestration to help operators advance toward autonomy.
And as part of GSMA’s new Open Telco AI initiative — launching tomorrow — NVIDIA is releasing the new open source LTM, implementation guide and agentic AI blueprints as open resources through GSMA, an organization for the mobile communications industry.
Open Nemotron 3 Large Telco Model Brings Reasoning to Telecom
For telcos to successfully operationalize generative and agentic AI across their operations, AI models must have the ability to understand the language of telecom and reason through complex workflows. NVIDIA has collaborated with AdaptKey AI to release a new open source, 30-billion-parameter NVIDIA Nemotron LTM that operators around the world can use to build autonomous networks.
Built on the NVIDIA Nemotron 3 family of foundation models and fine-tuned by AdaptKey AI using open telecom datasets including industry standards and synthetic logs, the LTM is optimized to understand telecom industry terminology and reason through workflows such as fault isolation, remediation planning and change validation.
As an open model, the Nemotron LTM gives telcos full transparency into how it was trained and what data was used, enabling secure and fast on‑premises deployment within their networks, where they can build and run agents directly. It also lets telcos safely adapt and extend telecom‑tuned reasoning with their own network and operational data, so they can move toward autonomous operations without sacrificing control over data or security.
Teaching AI Agents to Reason Like Network Engineers
NVIDIA and Tech Mahindra have published an open source guide that shows telecom operators how to fine-tune domain-specific reasoning models and build agents that can safely execute network operations center (NOC) workflows.
The guide outlines a framework for teaching models to reason like NOC engineers: focus on high‑impact, high‑frequency incident categories, translate expert resolutions into step‑by‑step procedures and turn those into structured reasoning traces that capture each action, tool call, outcome and decision. These traces become the “thinking examples” the model learns from, so it understands not just what to do, but why a particular sequence of checks and fixes is safe and effective.
Using the NVIDIA NeMo-Skills pipeline, operators can fine-tune a reasoning model on these traces, laying the foundation for telco-specialized AI agents that can reason and solve problems like a network engineer.
Maximizing Energy Efficiency With New Intent-Driven Energy Saving Blueprint
Autonomous networks rely on closed‑loop operation: models that understand the network, agents that act on intent and simulation that feeds results back into the system to validate and refine decisions. The new NVIDIA Blueprint for intent-driven RAN energy efficiency brings these pieces together, helping operators systematically reduce power consumption in 5G radio access networks (RAN) while maintaining quality of service.
The blueprint integrates network test and measurement leader VIAVI’s TeraVM AI RAN Scenario Generator (AI RSG) platform to generate synthetic network data — including cell utilization, user throughput and other traffic patterns — and convert it into a simple, queryable format.
An energy planning agent then reasons over the synthetic data to generate energy-saving policies that can be simulated in AI RSG, allowing operators to safely validate energy-saving policies in a closed loop to meet their intent without changing live configurations or impacting subscribers.
Telcos Put the NVIDIA Blueprint for Network Configuration to Work
Cassava Technologies is using the blueprint to build Cassava Autonomous Network, an agentic platform designed to optimize Africa’s diverse, multi-vendor mobile network environment. The platform implements three agents: one to monitor the network and recommend configuration changes, one to apply changes with documentation and governance, and one to assess the impact of changes made and safely roll them back if they have unintended effects.
NTT DATA is implementing the blueprint to bring intelligence to traffic regulation, helping the network manage surges when users reconnect after an outage, and is deploying it with a tier 1 operator in Japan.
An AI agent looks at real-time demand across the network and then decides when and how to admit new users on specific cells. As conditions stabilize, the agent adapts its decisions, turning what used to be manual configurations into a data-driven optimization cycle for more resilient mobile networks.
Evolving Network Configuration With Multi-Agent Orchestration
BubbleRAN is integrating NAT and BAT into its Opti-Sphere platform to manage network monitoring, configuration and validation agents more flexibly across containers and workloads, and connect them to tools that report network metrics and traffic status so they can continuously propose and validate configuration changes.
Telenor Group will be the first telco to adopt the blueprint with BubbleRAN to enhance its 5G network for Telenor Maritime, the group’s global connectivity provider at sea.
Learn more about the latest advancements in agentic AI for telecommunications at Mobile World Congress, taking place in Barcelona from March 2-5.
See notice regarding software product information.
Ransomware attackers are increasingly employing sophisticated tactics designed to defeat traditional defenses. Relying on basic security is no longer enough to protect your organization’s data. This checklist provides a comprehensive framework to help you develop an effective protection plan, covering everything from email entry points to lateral movement prevention.
In this checklist, you will learn how to:
Protect Your Email: Implement AI-enabled phishing protection and establish ongoing user training to secure your primary entry point.
Secure Your Applications: Utilize Zero-Trust access and network segmentation to prevent attackers from moving laterally through your network.
Fortify Your Backups: Learn why attackers target your backups first and how to protect them using encryption and IP restrictions.
Develop a Recovery Plan: Move beyond technical fixes to establish a full business response plan, including forensic analysis and full-scale testing.
Master the Basics: Close common gaps with updated patching schedules, strong password management, and multi-factor authentication (MFA).
Lenovo has a long-standing habit of showing up to MWC with at least one laptop concept that makes you stop in your tracks and double back for a closer look. Usually, those concepts are eye-catching but clearly something that makes you go – “cool, but never happening.” The ThinkBook Transparent Display Laptop we saw back in 2024 was one such futuristic laptop, but at MWC 2026, things felt different.
It’s not the kind of thing you just admire behind glass; it’s actually something you can see yourself using.
Lenovo showcased its new ThinkBook Modular AI PC concept, and surprisingly, it felt very normal. Don’t get me wrong; this is a good thing. This means the laptop is not the kind of thing you just admire behind glass; it’s actually something you can see yourself using in real life, albeit with some minor tweaks and refinements.
From the front, the ThinkBook Modular AI PC concept looks like any other 14-inch touchscreen business laptop. It’s thin, clean, and professional. There’s nothing about it that instantly screams “concept.” But then you walk around to the back, and that’s where the real story slowly starts to reveal itself.
Firmly attached to the rear lid via pogo pins is a second 14-inch display. It sits flush, almost invisible until you actually see it, and one use case that comes to mind almost immediately is sharing your screen with someone sitting across from you. Instead of awkwardly swiveling your laptop or having people huddle around your screen, you can just have one screen facing you, while the other faces the person across the table. Neat, right? But that’s not all…not by a long shot.
Don’t want to miss the best from Android Authority?
That secondary display at the back isn’t fixed. In fact, it’s magnetic and detachable, and Lenovo clearly designed it to be more than a gimmick. Pop it off, and you have a whole other portable display that can be mounted using a kickstand. You can use the secondary screen both vertically and horizontally, though it does need power from a separate pogo pin adapter.
Adamya Sharma / Android Authority
Things get even more interesting once you start rearranging the core pieces. The keyboard itself is removable, and you can swap it out for that second display, using the keyboard wirelessly with what’s now a dual-screen workstation.
The way all of this just works didn’t feel like I was using a concept laptop; it felt like I was using a laptop that just happened to be way more flexible than usual.
Then there are the brilliantly removable and swappable ports. The IO modules can physically shift from left to right (and vice versa) depending on how your desk or setup looks that day. USB-A, USB-C, and HDMI are all modular and move for your convenience. There’s even a small case that stores the removable ports, so you don’t lose them once they are out of the laptop.
The ThinkBook Modular AI PC feels more like Lenovo is testing what it should do next.
Under the hood, the ThinkBook Modular AI PC gets an Intel Core Ultra 7 255H processor, 32GB memory, and 1TB of M.2 PCIe SSD storage. There’s a 33Wh battery powering the whole thing, which is quite unfortunate, but then again, the story here isn’t the specs. This is a concept laptop, and one that Lenovo can always refine and upgrade if it ever becomes a reality. We may never see the whole laptop release as is, but parts of its modular design could very well show up in Lenovo’s retail units in the future.
Lenovo has shown ambitious modular ideas before, and several of them have quietly turned into shipping products over the years. Similarly, the ThinkBook Modular AI PC doesn’t feel like Lenovo is showing off what it can do; it feels more like Lenovo is testing what it should do next.
Thank you for being part of our community. Read our Comment Policy before posting.
Nested Lands is a ruthless medieval RPG game featuring in-depth village building, large-scale social management and combats. Gather survivors, build a settlement, craft, master your skills, survive, and endure alone or in online co-op.
Nested Lands is a multiplayer settlement-building RPG set in a dark, plague-ravaged world where twisted madmen and bandits roam the land. Your life will depend on how well you manage and grow your settlement. Fortunately, you’re not alone in this dark world — your peasants will help you gather resources, craft items, and build defenses as you lead them through danger and privation. But remember, every choice you make has consequences and can lead your village to prosperity or a senseless death from the plague.
BUILD YOUR SETTLEMENT
Start from a clean slate and build a brand-new civilization. Assign each peasant a specific role and order them to build, craft, or gather the resources you need to survive the hardships you will face. With dozens of upgradable buildings and hundreds of items to craft and customize, including farms and animals, you’re free to shape your settlement as you see fit. Call of Duty: Black Ops 6
RECRUIT YOUR ALLIES
The only way to break through the endless darkness of Nested Lands is to gather the other survivors who share your plight, but have yet to fully give up hope. Save and recruit peasants during your journey across the world. But keep in mind that they will need your protection.
MAKE YOUR NAME
Start from nothing. Rise from the bottom, overcome monumental challenges, and make your fame echo across the Nested Lands. But be careful – all your decisions will be carefully watched and evaluated. Will your name be praised by all, or will its very sound instill fear into the common folk who dwell in this open world? The choice is yours.
SURVIVE IN A POST-PLAGUE WORLD
The disfigured nature of the world makes survival even more challenging. Before facing other threats, you need to secure the basics — a lack of medicine can wipe your village out faster than any blade. Craft, improve your skills, and endure the harsh reality of Nested Lands. Brutal weather and the coming winter may be your greatest foes, but random encounters won’t offer any respite.
Features and System Requirements:
Nested Lands is a dark survival city-builder set in a plague-ravaged medieval world.
Establish and grow a settlement while defending it from raids, disasters, and deadly diseases.
Manage resources carefully as hunger, fear, and sickness threaten your people’s survival.
Build walls, train defenders, and make hard moral choices to keep hope alive.
Its grim atmosphere and challenging systems create a tense, immersive survival experience.
Screenshots
SystemRequirements
Minimum
OS *: Windows 10 and above
Processor: Intel(R) Core(TM) i3 or higher
Memory: 16 GB RAM
Graphics: NVIDIA GeForce RTX 3050 4GB or higher
DirectX: Version 12
Storage: 15 GB available space
Support the game developers by purchasing the game on Steam
InstallationGuide
TurnOff Your Antivirus Before Installing Any Game
1 :: Download Game 2 :: Extract Game 3 :: Launch The Game 4 :: Have Fun 🙂
Generative AI has been showing up in games so often these days that I don’t know who to trust anymore. It seems like every day there’s another game studio or publisher plaguing their games with more gen AI voice acting, gen AI chat bots, or gen AI art.
Blizzard developers, however, have come out against gen AI, implying it’s not being used in the development of their games. Overwatch game director Aaron Keller told GameSpot last month that the team isn’t “comfortable” putting “AI-generated content out in front of players.” Former World of Warcraft franchise director John Hight explicitly said no gen AI was being used for the MMO last year, too.
Speaking to Game Informer in a recent interview, WoW: Midnight’s lead composer Leo Kaliski says gen AI isn’t being used to make the score either. “I think we here at Blizzard feel very lucky and happy that we’re not using generative AI,” he says. “We’re just writing what we think is cool, and are happy to do that.”
Kaliski’s comment almost makes it sound like all of Blizzard isn’t using gen AI, which could be true. But given the specificity of the past comments from each team and Blizzard president Johanna Faries’ recent statement to Eurogamer about how the company has a “centralized governance team specifically focused on AI and the role it can or should play in development cycles,” I want to be careful assuming too much here.
That said, I think it’s notable that Blizzard doesn’t seem to be trying to control its messaging around gen AI use, and that it might be allowing the teams to decide if they want to experiment with it or not. This is much different than a CEO who picks up a megaphone and shouts about the glory of AI and how it alone will pioneer the company’s future.
Unfortunately, a lot of this could change at any point, especially when we know just how much Blizzard’s parent company, Microsoft, is determined to inject gen AI into everything it creates. The best we can hope for is that the developers on the inside can keep it away from their games. I wouldn’t be surprised if the relatively new unions at Blizzard are trying to fight this wherever they can.
Keep up to date with the most important stories and the best deals, as picked by the PC Gamer team.