Good Morning! Unity has announced official support for Apple's new Vision Pro headset, allowing developers to leverage Unity's existing XR platform to build immersive apps and experiences. Over 25,000 tech workers have already lost their jobs in the first few weeks of 2024, as companies like Meta, Amazon, and Microsoft cite overhiring and weak demand to justify layoffs that experts attribute more to herd behavior. Codeium, a startup providing an AI toolkit to boost developer productivity, has raised $65 million in Series B funding to expand its platform across the software development lifecycle.

— Forrest Knight & Meghanadh Vasireddy

Unity Launches XR Development Tools for Apple's Vision Pro Headset

Unity has announced official support for building augmented and virtual reality experiences on Apple's new Vision Pro spatial computing headset. This allows developers to leverage Unity's existing XR development platform to build immersive apps for Vision Pro's advanced capabilities.

Unity is providing project templates, code samples, and documentation to help developers get started with the Vision Pro SDK. This includes integration with Unity's PolySpatial technology for blending digital content with physical spaces. The goal is to enable a smooth transition for Unity devs to build Vision Pro apps using familiar tools and workflows.

Early adopters like Light Brick Studios and Resolution Games are already creating Vision Pro launch titles with Unity. Areas like automotive, meditation, and gaming will be represented. Unity is also the first major third-party platform to announce Vision Pro support, addressing concerns over lack of apps for the new device.

The Vision Pro support is part of Unity 2022 LTS, available now to Pro, Enterprise, and Industry tier subscribers. For XR developers accustomed to Unity, this is welcome news for easily targeting Apple's advanced stereo pass-through AR when the Vision Pro headset ships February 2nd, 2024.

Read More Here

25,000 Tech Workers were laid off in the first weeks of 2024.

In the first few weeks of January, around 25,000 tech workers lost their jobs across nearly 100 companies like Meta, Amazon, and Microsoft.

While execs cite pandemic overhiring and weak demand to justify last year’s 260,000+ job cuts, the same rationale seems dubious now. Inflation has fallen significantly, consumer confidence is rebounding, and most tech giants have returned to pre-covid staffing levels. Yet companies like Alphabet, TikTok, and Salesforce proceeded with additional layoffs.

So what explains this ongoing downsizing trend? Experts point to "herding behavior" and "copycat layoffs" in the tech industry. Essentially, companies mimic their competitors’ workforce cuts, perhaps to impress shareholders. The data supports this thesis - tech stocks like Microsoft have repeatedly hit all-time highs amid the recent layoff announcements.

Additionally, with interest rates rising, tech firms aim to cut costs and refocus investments towards AI advances like generative models. However, most industry titans remain enormously profitable, so survival hardly necessitates more layoffs.

As professors Jeffrey Pfeffer and Roger Lee explain, workforce cuts spread through the tech industry like a “self-fulfilling prophecy.” Ultimately, most data suggests tech layoffs result more from social contagion than economic hardship. Unless the culture changes, we can expect more of the same in 2024 and beyond.

Read More Here

Codeium raises $65 million in Series B funding for AI tools for developers

Codeium, a startup providing an AI toolkit to boost developer productivity, has raised $65 million in Series B funding. The round was led by Kleiner Perkins, with participation from previous investors General Catalyst and Greenoaks.

Codeium will use the new funding to expand its platform to automate more parts of the development lifecycle like system design, code maintenance, and continuous scanning. The company also plans to grow its engineering and sales teams.

Codeium's toolkit leverages AI to help developers write, analyze, debug, and deploy code more efficiently. Key capabilities include autocomplete, chat, and context-aware recommendations across over 70 programming languages. The platform already integrates with 40+ IDEs and is used by 300,000+ developers today.

As Varun Mohan, Codeium's CEO explains: "Software engineering is critical for innovation yet the process remains expensive and inefficient. With the rise of large language models, developers now have an opportunity to overcome these challenges and boost productivity."

Read More Here

Princeton researchers say generative AI isn't replacing devs any time soon

A recent study submitted to the International Conference on Learning Representations evaluates how well large language models (LLMs) can solve common software engineering problems from GitHub issues. Researchers from Princeton and the University of Chicago tested models including Claude and GPT-4 on nearly 2,300 real-world bugs and feature requests.

The results show current LLMs still struggle with many basics of software development. When provided with issue descriptions and corresponding code, the best model could only correctly solve 4.8% of problems. Researchers say isolating bugs from large codebases and reasoning about complex interdependencies across files remain difficult for AI.

The specialized model SWE-Llama, tailored for software tasks, marginally outperformed Claude and GPT-4. But its 4% success rate highlights the gap between industry excitement over coders' assistants and their ability to engineer solutions independently. Study co-author Carlos Jimenez says tools that simplify aspects of development will continue improving, but fully automated programming is still far off. For now, developers’ unique strengths in strategic reasoning and systems thinking remain indispensable.

Read More Here

Youtube Spotlight

I Spent 24 Hours Wearing Apple’s Vision Pro Headset | WSJ

The video describes the Joanna Stern’s 24-hour experience wearing Apple’s Vision Pro headset, testing its functionality and features in various scenarios including work, cooking, watching movies, and skiing, as well as highlighting its limitations and potential future applications.

Was this forwarded to you? Sign Up Here