One Of May’s Free PS Plus Games Just Leaked Online And It’s Scary Good

We’ve had very few PlayStation Plus leaks lately, but it appears we may know what at least one of next month’s PS Plus Essential titles is going to be.
PS Plus Essential titles are released on the first Tuesday of the month, with a reveal of the lineup the previous Wednesday, which means we won’t know for sure what’s coming next month until the last day of April. But this seems pretty likely given the timing.
As spotted by reddit user ShaneTVZ, it appears the Until Dawn remake could be one of May’s free titles. The reddit user posted this screenshot on the PlayStation Plus subreddit:
Until Dawn originally released way back in 2015. It remains one of my favorite horror games, taking the classic “group of attractive young people goes on vacation and gets killed one-by-one” movie trope and making it an interactive adventure where players have to make choices and outcomes can vary. It’s been years since I played, but if the 2024 remake is one of next month’s PS Plus offerings, I’ll happily dive back in.
The game has quite the cast. Hayden Panettiere, Peter Stormare, Rami Malek and lots of other talented actors offer their voice acting skills and likenesses, making this a genuinely cinematic experience.
The original game was developed by Supermassive Games, while the remake was handled by Ballistic Moon. The remake itself did not fare well with critics, though it has a “Mostly Positive” rating on Steam. The original game didn’t come to PC, so this was the first time PC players were able to take it for a spin. Most of the criticism from gamers is lackluster performance. Critics mostly said the remake felt more like a remaster and was largely unnecessary, adding little value to the original with only mildly improved graphics.
Forbes Daily: Join over 1 million Forbes Daily subscribers and get our best stories, exclusive reporting and essential analysis of the day’s news in your inbox every weekday.
Email address
Sign Up
By signing up, you agree to our Terms of Service, and you acknowledge our Privacy Statement. Forbes is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.
Still, free is free and I’d happily dive back in now that I’ve had a decade to forget the story. The leak seems especially likely given that the Until Dawn movie just hit theaters and appears to be a hot mess. I’m not surprised it’s getting poor reviews. When the trailer first released, the only thing I could think was how little it actually looked like the game it was purportedly based off of, which is weird given how easily this game could translate to the big screen.
What the other two (or possibly three) PS Plus games headed to May’s lineup are remains a mystery. We’ll know more this Wednesday. And obviously, since this is not confirmed, take everything with a grain of salt.
Powering Generative AI With The LLM Engine
The rapid advancement of generative AI (GenAI) is fundamentally reshaping the modern workplace, creating new businesses and solutions unlike anything imaginable just two years ago.
Large language models (LLMs) emerged in 2022 as a technological breakthrough comparable to the invention of the internal combustion engine in the late 1800s. Just as the engine converts raw fuel into power, LLMs transform data into intelligent, generative content that can automate complex tasks, boost productivity, optimize processes and enhance customer experiences.
But just as the world’s most powerful racing engine can’t win a race without a precisely engineered chassis, drive trains, and control systems, LLMs too need a robust framework of supporting technologies to create full GenAI applications that deliver business value with performance, safety, reliability and cost effectiveness. For GenAI applications, the LLM needs to be supported with components like data platforms, training or fine-tuning capabilities, embedding and vector databases to do retrieval augmented generation (RAG), monitoring and guardrails for safety and performance, and deployment and change control capabilities.
Selecting The Right LLM For Your Needs
Soon after the invention of the engine and its use to power vehicles, there was a rapid proliferation of vehicle types, with a range of special purpose engines, with matching chassis, drivetrains and control systems. There is no one engine that can power every vehicle type. A Formula 1 car’s powerful engine excels on the racetrack, but would fail as a delivery vehicle.
With LLMs, the same is true. Most providers have a range of model variants, from flagship ones with high levels of inherent knowledge and reasoning, to a range of smaller, faster, cheaper ones for simpler tasks like summarization that don’t require knowledge or reasoning. Model providers are also introducing specialized models trained on specific sets of data like legal, medical, financial, regional languages and translation, customer service, etc., to power industry specific use cases. More and more organizations are choosing to fine-tune third-party models with their proprietary data to improve performance in their use cases.
The evolution of LLMs continues to expand these possibilities, with newer models offering enhanced capabilities in processing text, images and videos. Another exciting development is the use of multiple models in a network of agents to achieve more complex outcomes with greater levels of automation.
While new LLM announcements appear weekly with impressive benchmark scores, focusing solely on leaderboard positions isn’t the most effective approach for your business. Successful organizations start by understanding their specific customer needs and use cases. You can then select an LLM from established providers that best matches your requirements for accuracy, speed, and cost-effectiveness. By building your solution with well-defined APIs, you maintain the flexibility to evaluate and integrate newer models as they become available. This approach helps you deliver consistent value while taking advantage of rapid innovations in LLM technology.
Protection Systems For GenAI Solutions
To build trusted GenAI applications, you need multiple layers of protection working together—similar to how modern vehicles combine various safety systems, from structural safeguards to advanced collision prevention.
Think of your application’s foundation layer as its structural framework. Just as a vehicle needs a robust chassis and engine mounting system, your GenAI application requires essential operational components. These include interfaces for model selection and interaction, systems to manage tokens and API calls, and mechanisms for handling prompts and responses. Your foundation also needs memory management capabilities, performance optimization through caching, load balancing for stability, and basic error handling protocols to ensure reliable operation.
Building on this foundation, you need active protection mechanisms—similar to how vehicles use seat-belts and collision detection systems. These protection systems safeguard your application through content moderation, input validation, and output verification protocols. Like a modern vehicle’s advanced driver assistance systems, these controls actively monitor operations, detect potential issues, and prevent harmful outputs. This layer includes governance policies, bias detection, content filtering, and comprehensive audit logging—all working together to ensure your GenAI application operates safely and reliably within defined parameters.
Your protection requirements vary based on your specific use case. For customer service applications handling sensitive data, you need comprehensive content filtering, strict input validation, and thorough output verification—comparable to the multiple security systems in an armored vehicle. Internal document processing applications might require basic content controls and standard validation, similar to the essential safety features in a delivery vehicle. When processing proprietary data in enterprise applications, you need strict access controls and comprehensive audit logging, much like the monitoring systems used in commercial fleet vehicles.
When you build these protection systems into your GenAI application from the start, you create solutions that are not only powerful but also trustworthy and reliable in production environments. Remember, the reliability and safety of your GenAI application are just as critical as its performance capabilities.
Building Your Complete GenAI Solution
After defining your business use case, identifying suitable LLMs, and planning your supporting components and controls, you are ready to build your GenAI solution. AWS has a three-layer architecture that adapts to your specific requirements to help you move quickly from your vision to reality. This is similar to how automakers offer different levels of customization, from basic components to fully assembled vehicles. The foundation layer provides the essential building blocks: compute power, storage capabilities, custom silicon, and specialized data stores for training and running LLMs. This layer serves organizations that need to build or fine-tune their own models, much like automotive manufacturers that design and build their own engines. Leading organizations such as Anthropic use these foundational AWS components to power their LLM operations, while companies like Thomson Reuters, Intuit, and Pfizer use them to create specialized solutions.
For most organizations, building a custom LLM isn’t necessary—you can use existing models, each with different capabilities and cost considerations. That’s where Amazon Bedrock comes in. Think of Bedrock as your complete vehicle assembly platform, providing access to a variety of proven LLMs through a single, secure interface. You can evaluate different models to find the best fit for your use case and start building appl