Indian software service companies are facing changes in their contractual work due to client demands for increased productivity from generative AI (GenAI) and global economic uncertainties. Experts warn these changes could reduce profit margins for these firms.

Infosys, Tata Consultancy Services (TCS), and HCLTech have all noted shifts in client contracts over the past two quarters, with clients expecting the same work for lower prices or expanded work scopes during contract renewals. This shift has put pressure on their profit margins.

HCLTech reported a 1.5% decline in IT services revenue in Q1 due to the end of major deals in financial services. The company's earnings before interest and tax (EBIT) margin also suffered as it transferred productivity gains to clients.

TCS CEO K. Krithivasan mentioned that clients are now adding more work scope during renewals to keep revenues stable, while Peter Bendor Samuel, CEO of Everest Group, highlighted that clients are pushing for greater productivity from GenAI, which may exceed current capabilities and pose profit risks.

Analysts from Nomura and Kotak Institutional Equities have observed similar trends, noting that productivity gains shared with clients are affecting margins. Nomura specifically pointed out a 50 basis point contraction in HCLTech's EBIT margin due to these productivity concessions.

Pareekh Jain, CEO of EIIRTrend, explained that clients' expectations have risen with the advent of GenAI, adding to traditional productivity methods like process improvement, automation, offshoring, and analytics.

Infosys also experienced contract renegotiations in the financial services sector, resulting in a slight revenue impact in Q4. Although some work was reduced, the majority of the contract remained intact.

OpenAI, the company behind ChatGPT, is developing a new project called "Strawberry" to enhance the reasoning abilities of its AI models. This information comes from an insider and documents reviewed by Reuters. The project is still in progress and its launch date remains unclear.

Strawberry is designed to help OpenAI’s models not just answer questions, but also autonomously navigate the internet to conduct what OpenAI calls "deep research." This level of advanced reasoning is something current AI models struggle with.

According to an internal OpenAI document from May, the company aims for Strawberry to significantly improve its AI’s ability to understand and plan tasks, much like humans. However, the specifics of how Strawberry works are kept confidential, even within the company.

A spokesperson from OpenAI confirmed that continuous research is key to improving AI, but did not provide specific details about Strawberry.

Previously known as Q*, Strawberry has shown promise in internal demos by answering complex science and math questions that current models can't handle. During a recent company meeting, OpenAI demonstrated a project with new reasoning skills, though it's unclear if this was Strawberry.

Strawberry involves a special method of processing AI models after they've been initially trained on large datasets. This post-training, similar to a technique called "Self-Taught Reasoner" (STaR) developed at Stanford, allows models to improve their intelligence by generating their own training data.

One of Strawberry’s goals is to handle long-term tasks that require planning and multiple steps. OpenAI is testing these capabilities with a "deep-research" dataset and aims to have its models perform tasks autonomously on the internet, guided by a "computer-using agent."

This project is part of OpenAI's broader effort to advance AI reasoning, a critical step toward achieving human or super-human level intelligence. Other companies like Google, Meta, and Microsoft are also working on similar advancements.

As part of the National Supercomputing Mission (NSM), the Telangana government is asking the Ministry of Electronics and Information Technology (MeitY) to set up a supercomputing facility with a 20 petaflop capacity in Hyderabad's projected Artificial Intelligence (AI) metropolis.

Supercomputer performance is measured in petaflops, which are units of one quadrillion (or 20,000 trillion) floating point operations per second. The NSM plans to build and operate 24 supercomputers with a total power of more than 64 petaflops throughout India.

Telangana's IT secretary, Jayesh Ranjan, said, "We have asked MeitY to set up this facility in Hyderabad's AI city. We are willing to donate the land required for this use. The amount of processing capacity available directly affects how far AI can advance. The Development Center of Advanced Computing (CDAC), a body under MeitY, has expressed interest in the project.

With a total capacity of more than 20 petaflops, CDAC has already deployed 11 systems under the first two stages of NSM at a number of institutions, including IISc, IITs, IISER Pune, JNCASR, NABI-Mohali, and CDAC itself. Magesh Ethirajan, the director general of CDAC, was not available for comment.

Telangana AI Mission (T-AIM) and CDAC inked a Memorandum of Understanding (MoU) in April of last year to give 142 AI entrepreneurs backed by T-AIM's Revv Up accelerator program access to reasonably priced high-performance AI computing. For a brief while, these businesses may utilize free access to India's most potent AI supercomputing facility, which offers 410 AI petaflops. These capabilities are available to startups for around one-third the price of private cloud services.

In line with the IndiaAI Mission, Ranjan stressed that Telangana's IT ministry will formally ask the union IT ministry for assistance in establishing the AI city. He said, "We will be meeting with ministers from the Government of India in Delhi to discuss this plan."

About 30 to 50 percent of the 200 acres that make up the proposed AI city would be left aside for businesses to establish AI centers. The remaining acreage will be developed into commercial spaces and computing facilities, which will be rented out to AI-focused businesses in exchange for internet, electricity, and rent subsidies.

Talks on establishing data centers in the AI city have also started with private businesses like Yotta, a data center service provider affiliated with the Hiranandani Group, and Intel, which runs more than 100,000 servers worldwide. Ranjan clarified, "We will permit them to set up their data centers for AI computing, including hosting graphics processing units (GPUs). "An R&D and incubation facility for agencies interested in AI research, development, and innovation is something else we plan to set up."

The AI city will also have a skilling center that will offer training for fundamental AI activities like data purification and annotation, as well as PhD-level education in AI. "Our goal is to create an all-encompassing environment for the advancement of AI," stated Ranjan.

Angular recently launched Angular 18, which brings several new features, including zoneless change detection, a new developer hub, and improved server-side rendering. This version focuses on stabilizing new APIs, responding to developer requests, and enhancing the developer experience.

Angular 18 now supports zoneless change detection experimentally, removing the need for zone.js. This aims to improve performance by reducing change detection cycles and providing clearer stack traces. Developers can try this new feature by adding a provider to their application bootstrap:

bootstrapApplication(App, {
  providers: [
    provideExperimentalZonelessChangeDetection()
  ]
});

Alex Rickabaugh, a senior software engineer at Google, expressed the importance of zoneless support in Angular 18, emphasizing the balance between maintaining stability and evolving with the modern web.

The official Angular documentation site, angular.dev, now includes a hands-on getting started guide, interactive playgrounds, updated guides, and easier navigation. All requests to angular.io now redirect to angular.dev.

Several features have become stable in Angular 18. Material 3 components are now stable, offering new themes and documentation. Deferrable views, which help improve Core Web Vitals, are now stable and allow developers to delay the loading of certain views. The new built-in control flow syntax also reaches stability, offering better performance and usability.

Server-side rendering (SSR) in Angular 18 is improved with i18n hydration support, enhanced debugging, and event replay using Google's event dispatch library. These improvements ensure a better and more interactive SSR experience.

Angular 18 allows developers to specify default content for ng-content, providing fallback content in their components. For example:

@Component({
  selector: 'app-profile',
  template: `
    Hello 
    Unknown user
  `,
})
export class Profile {}

Using the component:

<app-profile>
  <span class="greeting">Good morning </span>
</app-profile> 

<span class="greeting">Good morning </span>
Unknown user 

const nameControl = new FormControl('name', Validators.required);
nameControl.events.subscribe(event => {
  // Process individual events
});

const routes: Routes = [
  { path: "first-component", component: FirstComponent },
  {
    path: "old-user-page",
    redirectTo: ({ queryParams }) => {
      const errorHandler = inject(ErrorHandler);
      const userIdParam = queryParams['userId'];
      if (userIdParam !== undefined) {
        return `/user/${userIdParam}`;
      } else {
        errorHandler.handleError(new Error('Attempted navigation to user page without user ID.'));
        return `/not-found`;
      }
    },
  },
  { path: "user/:userId", component: OtherComponent },
];