TL;DR

• “Now and Then” by The Beatles is the first AI-assisted song nominated for Grammy awards, including Record of the Year and Best Rock Performance.
• AI technology, specifically “stem separation,” was crucial in isolating John Lennon’s vocals from a low-quality cassette, enabling the completion of the song.
• The song’s release in November 2023 generated global excitement and emotional responses from fans, highlighting AI’s role in preserving artistic legacies.

The Beatles Final Encore and Grammy History

In a remarkable combination of heritage and innovation, “Now and Then” by The Beatles has achieved the distinction of being the first AI-assisted song nominated for Grammy awards, signifying an important development in both the music industry and the influence of technology on creative fields.

Recognised for Record of the Year and Best Rock Performance, this track not only displays the enduring appeal of The Beatles but also emphasises AI’s ability to revive and maintain artistic legacies for future audiences.

The Background of “Now and Then”: Recovering a Forgotten Track

1. The Late-70s Demo and Lennon’s Heritage:

Originally an unfinished recording by John Lennon, “Now and Then” was never finalised due to technical difficulties. Captured on a low-quality cassette, Lennon’s vocals could not be separated from the piano backing, and his intentions for the song remained unclear, with just fragments of lyrics and melody guiding the band years later.

2. Abandonment and Resurgence Attempts:

In the mid-1990s, Paul McCartney and the rest of The Beatles tried to finish the song but eventually abandoned the project because of sound quality concerns and possible differences in artistic vision. With advancements in machine learning, filmmaker Peter Jackson revived this aspiration in 2021, employing sophisticated audio isolation techniques to extract Lennon’s vocals and breathe new life into the song. This technology, developed for Jackson’s Get Back documentary, proved essential in revealing the band’s original artistic intentions.

The Impact of AI on “Now and Then” as AI-assisted music

1. AI-Driven Audio Isolation:

The turning point emerged from an AI-enhanced method known as “stem separation,” a state-of-the-art technique designed to isolate different audio components. In “Now and Then”, the AI effectively separated Lennon’s voice from the piano, background sounds, and low-frequency noise. This process created a clear and distinct vocal track that could harmoniously integrate with recordings from McCartney, Starr, and Harrison’s archives.

2. Audio Restoration Focused on Enhancement:

Unlike generative AI, which creates new materials, the AI employed in this process concentrated solely on refining Lennon’s initial recordings. By enhancing and isolating the pre-existing audio, it kept Lennon’s genuine voice intact, honouring both the song’s legacy and its artistic authenticity. Producer Giles Martin emphasised the goal of enhancing Lennon’s voice without producing artificial elements, guaranteeing that the vocal delivery felt genuine and historically accurate.

3. AI Restoration Versus Traditional Methods:

Previously, in audio restorations, engineers depended on manual equalisation and noise reduction methods, which often compromised sound quality. In contrast, AI provided unmatched accuracy, striking a balance between fidelity and authenticity. The technology’s capacity to identify and extract complex audio layers enabled the production team to revive the song with a clarity that traditional techniques could not match.

The Last Beatles Recording: Creation and Release

1. Contributions from All Four Members of The Beatles:

“Now and Then” brings together all four members of The Beatles, creating a musical connection across generations. Along with Lennon’s vocals, McCartney’s bass, Starr’s drumming, and Harrison’s previously recorded guitar pieces merge into a final version that captures the essence of a true Beatles song. The contributions of each member were thoughtfully blended, resulting in a track that respects both their original spirit and their individual roles.

2. Premiere and Immediate Reaction:

Released in November 2023, “Now and Then” generated important global excitement, not only for its nostalgic value but also for the way it bridged the past and present through technology. It resonated as a heartfelt farewell, eliciting strong emotional responses from fans who found comfort in the song’s enduring melody and lyrics.

Grammy Nominations: A New Chapter for AI in Music

1. Establishing Grammy Milestones with AI:

The nominations for Record of the Year and Best Rock Performance signify the first instance that an AI-enhanced work has achieved this level of acknowledgement, opening doors for future AI-assisted projects. Competing against leading contemporary artists like Beyoncé, Taylor Swift, and Billie Eilish, The Beatles have returned to prominence, once more challenging creative limits even in Grammy qualifications.

2. Grammy Policies on AI and Human Creativity:

While the Recording Academy restricts the extent to which AI can be involved, this nomination is consistent with Grammy rules that allow AI to be used as a supportive tool rather than serving as the main creator. This nomination recognises that AI, when applied with care, can enhance human craftsmanship instead of replacing it, indicating new opportunities for the music industry as AI-supported work becomes increasingly prevalent.

Public and Critical Reaction: Emotional and Technological Impact

1. The Emotional Connection to Fans:

Numerous fans shared powerful emotional responses, recounting how “Now and Then” invoked memories of Lennon, Harrison, and earlier moments associated with The Beatles’ music. The song’s thoughtful nature generated a feeling of finality, moving listeners who sensed that they were receiving a last message from the band. Social media was filled with fans sharing their experiences of how the song brought forth strong memories and ties to their personal lives.

2. Musical Praise and Mixed Reactions:

Critics applauded the song’s evocative tune and production quality, although opinions were split regarding its contemporary production style. While some listeners enjoyed the daring modern sound, others felt it either leaned too much on McCartney’s influence or lost the authenticity that characterised earlier Beatles tracks. The inclusion of AI sparked a discussion over whether this piece could genuinely be classified as a Beatles song or if it strayed too close to a McCartney creation.

3. Ethical and Artistic Concerns:

Both critics and fans questioned the song’s genuineness, with some expressing concerns that AI-assisted tracks could undermine artistic integrity. The ethical conversation focused on whether the utilisation of AI in this way respects or distorts Lennon’s original vision, particularly considering Harrison’s disreported fondness for the song. Nevertheless, this debate also positioned “Now and Then” as a technological achievement, encouraging a conversation about AI’s capability to safeguard legacies.

Breaking New Ground in AI-Assisted Music: AI and The Beatles’ Legacy

1. An Authentic Technological Triumph:

The involvement of AI in “Now and Then” presents a future in which technology can assist rather than change musical craftsmanship. By assisting precise vocal separation, AI has enabled “Now and Then” to connect with audiences as a genuine Beatles song rather than as a synthetic production. The narrative surrounding this song also emphasises how AI can be used to preserve the authenticity of historical works, providing fresh insights into cherished artists.

2. Legacy, Nostalgia, and Artistic Responsibility:

As The Beatles make Grammy history with “Now and Then”, the song emphasises the details of employing AI within legacy music. This release demonstrates how, with thoughtful application, AI can rejuvenate memories without compromising authenticity, enriching The Beatles’ narrative in a manner that feels both innovative and reflective.

Looking Forward: Future Projects and Ethical Debates

1. Possibilities for More AI-Restored Works:

This release paves the way for future endeavours, where AI could potentially restore other unfinished tracks or archival content. However, as technology advances further, the ethical concerns raised by “Now and Then” will likely shape decisions about whether and how AI should be integrated into artists’ legacies.

2. AI in Modern Music Production:

The application of AI in “Now and Then” acts as a model for how AI could support the creation of new music while preserving artistic intention. Its success is expected to motivate other artists and estates to examine how technology can connect the past to the present, and what limits might be established to maintain authenticity throughout the process.

Old Rockers Meet New Technological Innovation with The Beatles

“Now and Then” stands as an important milestone, blending innovation with nostalgia and technology with tradition. In its pursuit of Grammy recognition, the song is a testament to the capability of AI to enrich human creativity, ushering in a new chapter in music history.

As The Beatles are honoured once again in February 2025, “Now and Then” could be recalled as both a homage to their legacy and an indicator of music’s AI-enhanced future.

You can listen to the music video on YouTube by tapping here

Join the Conversation

What are your thoughts on the future of AI in music and its potential to revive and maintain artistic legacies? We’d love to hear your opinions and experiences. Don’t forget to subscribe for updates on AI and AGI developments and share your comments below. Subscribe here.

You may also like:

• AI Music Generator: Creating Songs from Text Prompts
• AI Music Revolution: Suno Fights Back Against Record Labels
• How AI Solves the ‘Cocktail Party Problem’

TL;DR:

• AI is being explored as a tool in paranormal research, potentially bringing science to haunted house investigations.
• Machine learning could help detect patterns in spooky phenomena like temperature drops, strange sounds, and electromagnetic fluctuations.
• Tools like natural language processing and image recognition could assist in analysing ghostly voices (EVPs) and visual anomalies.
• AI might help explain away myths, but it could also lend credibility to genuinely unexplainable events.

Exploring the Unknown: How AI and Haunted Houses Are Becoming an Unlikely Pair

For centuries, haunted houses have captivated imaginations, igniting everything from spine-chilling horror tales to fervent ghost-hunting adventures. As Halloween brings the spectral into focus, the question lingers: can modern technology finally demystify haunted houses? Enter artificial intelligence – a tool more often associated with data and algorithms than with the supernatural.

Yet, as it turns out, AI may hold the keys to unlocking—or at least better understanding—paranormal occurrences. Imagine if data could explain creaks, shadows, and strange chills that have puzzled investigators for years. As improbable as it may sound, AI is already inching closer to solving the mysteries of the “haunted” world.

AI in Paranormal Research: A New Kind of Ghost Hunter

Until now, paranormal research has relied on a mix of psychology, folklore, and equipment that wouldn’t look out of place in a 1980s horror film. Ghost hunters use EMF readers, temperature gauges, voice recorders, and infrared cameras to collect evidence of supernatural phenomena. But what if these tools could be combined with AI to create a new level of precision?

AI excels at pattern recognition, so it’s perfectly suited to comb through hours of data, identifying correlations and potential anomalies that human eyes—or even traditional ghost-hunting gadgets—might miss. Machine learning algorithms could help sift through thousands of data points collected from haunted sites: sudden temperature drops, unexplained sounds, or electromagnetic fluctuations, each a piece of the larger puzzle. By systematically analysing these elements, AI could help us distinguish between mere environmental oddities and events that genuinely defy explanation.

The Tools of a Digital Paranormal Investigator

If AI is to become the new Sherlock Holmes of the spectral realm, it’ll need a few specialised tools. Let’s explore what the high-tech ghost-hunting kit might include:

Natural Language Processing for EVPs:

Electronic Voice Phenomena (EVPs) have long been considered potential evidence of the paranormal. Sceptics might argue that these “voices” are simply auditory pareidolia—our brain imposing meaning on random noise. But here’s where AI steps in. Using natural language processing (NLP), we could analyse EVPs more rigorously, identifying genuine speech patterns, tonal shifts, and distinguishing real words from background noise. With enough data, AI could start to build a profile of these sounds, either debunking them as misinterpretations or, intriguingly, revealing consistencies that suggest something else might be at play.

Image Recognition for Visual Anomalies:

Photographs are classic ghost-hunting artefacts, but they are frequently dismissed as lighting tricks, camera errors, or even hoaxes. However, machine learning algorithms trained on image recognition can analyse photographs, helping us separate artefacts from authentic phenomena. Suppose AI notices recurring shapes or unusual patterns in a particular location that can’t be explained by environmental factors or photo manipulation. In that case, it could give researchers a reason to look closer.

Anomaly Detection for Environmental Data:

Haunted houses are said to have their own unique “energy,” but until now, we’ve lacked a way to quantify it. Anomaly detection algorithms could monitor everything from temperature fluctuations to electromagnetic field (EMF) changes in real time, establishing baseline readings and alerting researchers when anything deviates. This way, AI isn’t just collecting data; it’s proactively looking for the moment when something out of the ordinary occurs. What’s more, it could correlate multiple data streams—such as a sudden cold spot occurring alongside a spike in EMF readings—shedding light on whether there’s any interaction between these phenomena.

Bridging the Rational and the Supernatural

AI isn’t just about debunking hauntings. In fact, it could provide believers and skeptics alike with a clearer, balanced perspective on what’s truly happening in these so-called haunted spaces. For every draft that causes a door to creak or faulty wiring that triggers strange lights, AI could offer a logical explanation, potentially demystifying many phenomena.

However, the beauty of AI is that it may also uncover genuinely unexplainable occurrences. Imagine a system that has accounted for every variable—temperature, humidity, pressure, electrical interference—and still records an anomaly that defies scientific understanding. In such instances, AI’s credibility could lend weight to claims of paranormal activity, especially in cases where all other factors have been ruled out. Far from rendering hauntings obsolete, AI could lend them a new form of credibility, giving credence to reports that have long been dismissed.

Treading Lightly on Ancient Beliefs

Using AI to delve into the paranormal isn’t merely a technical endeavour; it’s a cultural one as well. Haunted houses and ghost stories are not just frivolous folklore; they’re woven into the cultural fabric of many societies. Take Asia, where supernatural beliefs are strong, with ghosts and spirits playing roles in local traditions and societal norms. For some, these “haunted” places hold emotional significance, representing ancestors or spirits that are seen as part of everyday life rather than objects of fear.

This raises an ethical question: should AI be used to challenge or even debunk such beliefs? Technology has a long history of clashing with tradition, and there’s a risk that AI could be seen as a tool for eroding cultural heritage. An overly clinical approach could turn paranormal research into something that disregards the cultural importance of these stories. A respectful, balanced approach could instead enhance the understanding of paranormal phenomena without undermining long-standing beliefs. By respecting the human aspect of these tales, AI can be both a scientific ally and a cultural bridge.

Could AI Ever Truly Explain the Paranormal?

In the end, AI may bring a level of insight to ghost-hunting that was once the realm of imagination. By grounding supernatural research in data, AI could enable a new generation of paranormal investigators to investigate the unknown with greater precision, nuance, and respect. But could it ever truly explain the unexplainable? Probably not. Even with the most sophisticated technology, some elements of the paranormal may forever elude scientific understanding, preserving that essential bit of mystery that keeps haunted houses and ghost stories alive in our minds.

Comment and Share:

Perhaps, though, that’s for the best. In a world where data is constantly reducing unknowns, maybe we need a few things to remain beyond the grasp of AI. After all, what would Halloween be without the thrill of the unknown, the shiver down our spine, and the occasional chill in the air that no data set can quite explain? Subscribe for updates on AI and AGI developments and share your thoughts in the comments below!ou may also like:

You may also like:

• 7 GPT-4o Prompts That Will Blow Your Mind!
• 6 Stunning AI-Generated Visual Effects for Aspiring Artists
• Adrian’s Arena: AI is Shaping the Future of Marketing to Gen Z in Southeast Asia
• Now download some spooky designs this Halloween for free from website, Perfect

TL;DR:

• MarioVGG is a new AI model that can generate plausible video of Super Mario Bros. from user inputs.
• The model was trained on over 737,000 frames of Mario gameplay.
• Despite limitations, MarioVGG shows potential for AI to replace game engines in the future.

The Future of Gaming: AI-Generated Video

Imagine playing your favourite video game without a traditional game engine. Instead, an AI model generates the gameplay based on video footage. This is the fascinating concept behind MarioVGG, a new AI model that simulates Super Mario Bros. from video data. Developed by researchers from Virtuals Protocol, MarioVGG represents a significant step towards AI-generated video games.

Training MarioVGG: A Massive Undertaking

To train MarioVGG, the researchers used a public dataset containing 280 levels of Super Mario Bros. gameplay. This dataset included over 737,000 individual frames, which were preprocessed into 35-frame chunks. The model focused on two inputs: “run right” and “run right and jump.” Even with these limitations, training the model took about 48 hours on a single RTX 4090 graphics card.

How MarioVGG Works

MarioVGG uses a standard convolution and denoising process to generate new frames of video from a static starting game image and a text input. The model can create gameplay videos of any length by using the last frame of one sequence as the first frame of the next. This results in “coherent and consistent gameplay,” according to the researchers.

Challenges and Limitations

Despite its impressive capabilities, MarioVGG has several limitations. The model downscales the output frames to a resolution of 64×48, much lower than the NES’s 256×240 resolution. It also condenses 35 frames of video into just seven generated frames, resulting in rougher-looking gameplay. Additionally, MarioVGG struggles to approach real-time video generation, taking six seconds to generate a six-frame video sequence.

Impressive Results Despite Limitations

Even with these limitations, MarioVGG can create passably believable video of Mario running and jumping. The model can infer game physics, such as Mario falling when he runs off a cliff and halting his forward motion when adjacent to an obstacle. MarioVGG can also hallucinate new obstacles for Mario, although these can’t be influenced by user prompts.

The Future of AI in Gaming

The researchers hope that MarioVGG represents a first step towards “producing and demonstrating a reliable and controllable video game generator.” They even suggest that AI models like MarioVGG could one day replace game development and game engines completely.

Comment and Share:

What do you think about the future of AI in gaming? Could AI models like MarioVGG really replace traditional game engines? Share your thoughts and experiences in the comments below. Don’t forget to subscribe for updates on AI and AGI developments.

You may also like:

• Unleash Your Creativity: Building Retro Games with AI Tools
• Revolutionising Gaming: The AI and AGI Impact in Asia
• AI Showdown: Video Game Performers Strike for AI Protections
• To learn more about AI & gaming, tap here.