This episode explores how Subnet 85 Vidaio, led by Gareth Howells, is revolutionizing video processing on Bittensor through AI-driven upscaling and compression, promising significant cost savings and quality enhancements for enterprise clients.
Gareth's Extensive Background in the Video Industry
- The Ventura Labs podcast introduces Gareth Howells, owner of Subnet 85 Vidaio on the Bittensor network.
- Gareth outlines his over 20 years of experience in the video industry, a journey that began with DigiBeta tapes and evolved through digitization, file-based systems, and eventually cloud-based video workflows.
- His expertise covers diverse areas such as news, animation, music, and sports, with a consistent focus on the operational aspects: “moving it around, compressing it, storing it, and essentially getting it onto the screens,” rather than content creation itself.
- Actionable Insight: Gareth's profound operational experience in traditional video provides him with a solid understanding of real-world enterprise requirements, a critical asset for Vidaio's business-to-business (B2B) strategy.
Experience in Emerging Markets and the Genesis of Vidaio
- Gareth recounts his decade of experience working in emerging markets, primarily in West Africa.
- He was instrumental in starting Iroco TV, often called the “Netflix of Africa,” a VOD (Video on Demand) platform. VOD systems allow users to select and watch video content whenever they choose. This platform was tailored for Africa's mobile-first audience, addressing specific challenges like limited bandwidth and high data costs.
- Gareth states, “now with Vidaio, this is exactly what we're doing. We're solving those problems... that we've been trying to solve for the last 10 years.”
- The concept for Vidaio arose from the absence of suitable solutions in the open market, leading to the decision to build it on Bittensor. Bittensor is a decentralized network that uses its native token, TAO, to incentivize the creation and operation of AI models. Vidaio operates as a Subnet on Bittensor, which is a specialized network dedicated to a specific AI task—in this case, video processing.
- Strategic Implication: Vidaio's origin in solving tangible problems for emerging markets, where efficiency is crucial, positions it strongly for broader enterprise adoption where cost and bandwidth are also significant considerations.
Journey to Africa and Entry into Bittensor
- Gareth shares how he transitioned from working at Pokémon on media production (including dubbing, digitizing, and manually upscaling their extensive animation archive) to the African market. This shift occurred after a headhunter contacted him for the Iroco TV project.
- His introduction to Bittensor came through his friend Mog Adham, a recognized figure in the Bittensor community. Mog explained the concepts of AI crypto projects, incentive mechanisms, and subnets.
- This discussion ignited the idea for Vidaio. Gareth then pitched this concept to Mog, which led to its development and Mog's involvement as an investor and advisor.
- Speaker Analysis: Mog Adham's early guidance and investment underscore the significance of network connections within the Bittensor ecosystem for founders launching new subnets.
Why Bittensor is Ideal for Vidaio's Video Upscaling
- Gareth highlights “evolution” as the primary advantage of Bittensor for Vidaio. He points to the rapid progress from a basic miner to outperforming competitors in upscaling within a few months. Upscaling is the AI-driven process of increasing video resolution (e.g., from 2K to 4K) by intelligently generating additional pixel data.
- He emphasizes the unique power of Bittensor's structure: “to have 256 machine learning engineers working specifically on your problem... trying to outdo the next guy that just creates such a powerful output for your product.”
- Actionable Insight: The competitive, incentivized environment of Bittensor's mining community accelerates AI model development at a rate that is difficult for traditional corporate R&D to match. This offers a substantial competitive advantage to subnets like Vidaio.
AI in Video Processing vs. Traditional Methods
- Gareth explains that traditional video upscaling was a laborious, manual, frame-by-frame process, making it prohibitively expensive for most content.
- In contrast, AI models can predict and intelligently fill in missing pixel data when increasing resolution (e.g., from 2K to 4K), rendering the process faster, more efficient, and significantly cheaper.
- “What the machine learning can do is it can understand what should be in those pixels to create a smooth clear picture.”
- This technological advancement unlocks vast archives of older content (often in Standard Definition, or SD) for repurposing and monetization by content owners.
- Strategic Implication: AI-driven upscaling can unlock dormant value in existing media libraries, creating a substantial addressable market for Vidaio's services.
The Core Value: AI-Driven Video Compression
- Gareth identifies video compression as the principal value proposition of Vidaio, even more so than upscaling. Video compression reduces video file sizes and bitrates to minimize storage and bandwidth needs while preserving perceived visual quality.
- The objective is to drastically reduce file sizes and bitrates for content (like 4K video) while maintaining “perceived quality”—how the video appears to the human eye.
- This AI-driven approach differs from traditional mathematical compression methods, which can often result in blocky, low-quality images, especially in low-bandwidth scenarios.
- “Where we're going to get big enterprise clients, big costsaving for our clients and efficiencies is from the compression.”
- Key Statistic: Vidaio aims for significant savings, such as a potential 58% reduction in storage costs for large content owners. This also leads to reduced expenses for bandwidth, CDNs (Content Delivery Networks), telecommunication companies, and end-users (through lower data usage and less battery consumption). CDNs are distributed server networks that deliver content, including video, to users based on their geographic proximity, enhancing speed and reliability.
- Vidaio’s compression model, which they built in-house due to the lack of suitable open-source alternatives, is demonstrating “really, really incredible” results in early tests and benchmarks.
- Actionable Insight: Vidaio's focus on AI-enhanced compression directly addresses a critical challenge for any organization managing large volumes of video. The potential for substantial, quantifiable cost savings makes it a highly attractive proposition for enterprise clients.
Revenue Model and Target Clients
- Gareth clarifies that Vidaio's revenue model will be predominantly B2B, focusing on large platforms, content owners, telecommunication companies, and CDN providers.
- While a consumer-facing product for both upscaling and compression will be available, its primary purpose will be to showcase the technology rather than serve as a major revenue stream.
- “Enterprise clients using the subnet on big scales is really where we see the revenue for the subnet coming.”
- Strategic Implication: A distinct B2B focus enables Vidaio to concentrate its resources on securing high-value enterprise contracts, effectively leveraging Gareth's extensive industry network.
Addressing Video Privacy and Crypto Skepticism
- Gareth acknowledges that video privacy is a significant concern that Vidaio must address, particularly for unreleased content or situations involving non-disclosure agreements (NDAs).
- Potential solutions include using the subnet to identify top-performing miners who could then process sensitive content off-chain under specific agreements, or, more preferably, developing on-chain encryption methods.
- He anticipates some initial reluctance from mainstream clients due to the project's “AI crypto decentralized open-source” nature.
- “We need to prove that the security and the safety to the clients but I don't think that's once they see the results I think and the cost I think that's something that you know we can quickly get them on board.”
- Regarding skepticism towards crypto, Gareth attributes this to prevalent negative narratives about scams and market volatility, especially among those unfamiliar with the space. He believes his established industry relationships will be key in overcoming these perceptions.
- Actionable Insight: Investors should closely watch how Vidaio tackles enterprise-level security and privacy concerns, as this will be pivotal for mainstream adoption. Overcoming crypto skepticism will depend on demonstrating clear, tangible benefits (cost savings, quality improvements) and leveraging trusted industry connections.
Gareth's Positive Experience in the Bittensor Ecosystem
- Despite being relatively new to the crypto space, Gareth describes his experience with Bittensor as “immensely positive.” He praises the supportive and collaborative network of miners, validators, and other participants.
- He notes that while the incentive-driven nature of the system can lead to attempts at exploitation, Vidaio has also received valuable, honest feedback from miners, helping to proactively identify and address potential issues.
- “I feel very grateful that I found it... I haven't had to go through 10 negative experiences to find one positive. So, you know, I think I hit the jackpot first time.”
- Speaker Analysis: Gareth's perspective as a newcomer from traditional industry offers a compelling endorsement of the Bittensor community's collaborative strengths, which exist alongside its competitive incentive structures.
Incentive Mechanisms and Combating Gamification
- Gareth explains that Vidaio's validation process centers on “perceived quality”—how a video actually looks to the human eye.
- They employ two primary video verification metrics:
- VMAF (Video Multi-Method Assessment Fusion): A metric developed by Netflix, widely regarded as a gold standard for perceived quality. However, in an incentivized environment like Bittensor, Gareth found it can be “really easily gamed.”
- PI app (Perceptual Image Quality Assessment): Another metric that Vidaio has found to be “incredibly powerful” and “much more difficult to game.” The transcript refers to it as “PI app,” likely indicating a specific perceptual index or a similar tool designed for image quality assessment.
- Early on, some miners used “pretty clever tricks,” such as pre-processing steps, to artificially inflate their scores. Vidaio countered these tactics by benchmarking results with human eye checks and implementing measures to prevent cheating, crediting their lead developer, Medfill, for his work in this area.
- “We managed to work out how they were doing it... and we've put things in place to stop that.”
- Actionable Insight: The ongoing efforts by subnet operators to prevent gamification by miners are crucial for maintaining the integrity and quality of the AI models produced. Vidaio's dual-metric approach and constant vigilance are key to this process.
The Role of VMAF and PI app in Validation
- The host clarifies the distinct functions of the two metrics: PI app is used to score the perceived quality of the processed video, while VMAF serves to verify that the video returned by the miner is the same video that was originally sent for processing, not a different, albeit high-quality, video.
- Gareth confirms this: “What we use VMAF to do is to check that the files the frames match up from the original video to the upscale video.”
- Technical Clarity: This two-pronged validation system is vital. One metric (PI app) assesses the quality of the AI's output, while the other (VMAF) ensures the integrity and relevance of the content, preventing miners from submitting unrelated high-quality videos to earn rewards.
Influence of Jan Ozer and Working with Existing Technology
- Gareth recounts an early and influential meeting with Jan Ozer (transcribed as “Yan Ozair,” a respected expert in video compression).
- A key insight from this discussion was the strategic decision for Vidaio to focus on working with and enhancing existing video technologies and codecs (e.g., H.264, H.265, AV1), rather than attempting to create an entirely new AI video codec from scratch. A codec (coder-decoder) is the technology used to compress (encode) and decompress (decode) digital video data.
- “Everything we're doing is driven by hardware... it's all about encoders and decoders. So what can your mobile phone decode?”
- Developing a completely new AI codec would likely face a very long adoption cycle (potentially 10 years) for hardware manufacturers to support it, making it commercially unviable in the near term.
- Strategic Implication: Vidaio's pragmatic approach of improving existing, widely adopted video codecs ensures immediate market relevance and a practical path to adoption. This is a critical strategic choice that avoids reliance on uncertain future hardware developments.
Competitive Landscape and Vidaio's Edge
- Gareth acknowledges the presence of traditional competitors in the video processing space, such as Topaz Labs and Encoding.com, and notes that many large platforms also handle their own encoding and transcoding. (The transcript mentions “Fijian,” which might be a transcription error or a less common company).
- Despite this, he asserts, “none so far that we have come across have been able to replicate what we're doing.”
- Vidaio’s custom-built compression engine, developed internally because no existing open-source project met their specific needs, is yielding “incredible results” in early tests.
- “There's not a single open-source project for video compression that basically was fit for purpose. So we had to go away and build one.”
- He emphasizes that even if their initial base model weren't superior, the rapid evolutionary nature of development on Bittensor would quickly enable them to surpass competitors. Vidaio's upscaling models, for instance, began outperforming competitors within a month of launching their base miner.
- Actionable Insight: Vidaio's creation of a proprietary compression engine, combined with the accelerated AI development inherent in the Bittensor network, provides a strong competitive differentiation. Investors should look for upcoming benchmark releases comparing Vidaio's performance against established industry players.
The Selling Point to Major Platforms
- Gareth is confident that the demonstrable results of Vidaio's technology will be compelling when pitching to major platforms like Netflix or YouTube.
- A significant selling point is cost savings. The Bittensor structure allows Vidaio to operate without the substantial overhead costs of traditional R&D departments or expensive corporate offices. These savings can be passed on to their partners.
- “We'll be much cheaper than anyone else that is out there and we'll be much more efficient.”
- The strategy involves proving the concept and its benefits in emerging markets first, then expanding to Western markets as the technology's value becomes evident.
- Strategic Implication: The combination of superior technology (offering enhanced quality and efficiency) and a lower cost structure forms a powerful value proposition for large enterprises seeking to optimize their video operations.
Role of the Subnet Owner and Vidaio's Tokenomics
- Gareth describes his role as multifaceted, involving deep engagement in product design and strategy, alongside a strong focus on business development, marketing, and customer acquisition, leveraging his extensive industry network.
- “The product is nothing without a customer.”
- Regarding Vidaio's native subnet token (which the host refers to as the “alpha token,”), Gareth explains that its value is intrinsically linked to the anticipated success of the product in a vast market.
- Vidaio plans to reinvest revenues back into the subnet to fund further research, development (such as exploring streaming solutions), and overall ecosystem growth. This reinvestment strategy is designed to enhance the token's utility and value over time.
- Tokenomics: This refers to the economic model and design of a cryptocurrency, encompassing its supply, distribution, utility, and incentive structures. For Vidaio, holding its token signifies a belief in the subnet's future revenue generation and the utility it will provide.
- “We don't plan to... extract all of the revenue out of the subnet. We plan to, you know, put the money back in.”
- Actionable Insight: Vidaio's commitment to reinvesting profits into the subnet, rather than prioritizing immediate extraction, signals a long-term vision. This approach could significantly increase the token's utility and value as the platform matures and expands its offerings.
Impact of Video Compression on Internet Traffic
- The host underscores a critical point: video constitutes approximately 80% of all internet traffic. If Vidaio's technology can compress this video data by even 50%, it would lead to a dramatic reduction in overall internet traffic (e.g., effectively halving the bandwidth consumed by video).
- Gareth reiterates the immense cost savings this translates to across storage, bandwidth, and streaming for platforms.
- “If you can save them money in all areas, say we cut their storage cost by 50%... they're saving $50 million a year just by using our services and with no detriment to the quality of the product.”
- Strategic Implication: The sheer scale of potential cost savings positions Vidaio's compression technology as a highly disruptive force with the potential to reshape the economics of internet infrastructure and media delivery industries.
Future Plans: Decentralized Streaming
- Gareth points out that essentially all video processed by Vidaio is ultimately intended for streaming.
- Looking ahead, Vidaio is exploring the development of a “decentralized streaming platform” that would operate through Vidaio and the broader Bittensor network.
- This initiative could potentially leverage other Bittensor projects, such as “Hippius” (likely a decentralized storage or edge computing project mentioned in the transcript), to create a distributed infrastructure for streaming.
- “If we can utilize that technology utilize our technology to make the picture quality better to make the the bandwidth smaller we can have a really successful streaming project built within the bit tensor ecosystem.”
- Initially, this would likely take the form of a backend service—for example, a white-label application allowing podcasters to stream their content decentrally—rather than a full-fledged competitor to platforms like YouTube. Vidaio might partner with other decentralized platforms for front-end development and content curation.
- He also notes that another entity is reportedly launching an “AI Netflix” subnet, which could potentially become a user of Vidaio's advanced video processing technology.
- Emerging Trend: The development of decentralized streaming infrastructure within the Bittensor ecosystem represents a significant emerging trend. Vidaio aims to position itself as a core technology provider for such services, offering superior quality and efficiency.
Gareth's Focus and Post-DTA Strategy
- Gareth briefly mentions an early, unsuccessful attempt at mining on Bittensor, undertaken to better understand the system. However, he does not currently engage in trading; his entire focus is on developing Subnet 85 Vidaio.
- DTA (Decentralized TAO): The host's term “detail” throughout this segment likely refers to DTA. This term often signifies a phase in Bittensor's evolution, possibly related to broader token distribution or a more mature ecosystem state, where subnets face increased scrutiny and need clearer value propositions.
- He stresses the critical importance, especially in the current “post-DTA” Bittensor environment, of having a viable product and exceptionally clear messaging that the community and potential investors can easily grasp.
- “I hope that people can understand what we're doing really easily... I think it's important that I try and be as clear with our message, what we're doing, what the market is, what our successes are.”
- This clarity in communication is vital for discoverability and attracting investment in an increasingly crowded landscape of subnets.
- Actionable Insight: For investors evaluating subnets, clear communication, a demonstrable product, and a well-articulated market strategy are becoming increasingly important differentiators in the evolving Bittensor ecosystem.
Key Takeaways from the Bittensor Experience
- Gareth reiterates that the inherent power of the subnet model and its capacity for rapid evolution, driven by the contributions of miners and the incentive mechanism, are standout features of the Bittensor network.
- He expresses amazement at the speed with which it's possible to establish a potentially successful technology company and develop advanced products within the Bittensor framework—an achievement he believes would have been unattainable in the traditional, centralized world without protracted and extensive fundraising efforts.
- “The evolution and the adaptability of what we're doing is just incredible.”
- Strategic Implication: Bittensor offers a unique and powerful environment for fostering rapid innovation and facilitating market entry for AI-focused companies, significantly lowering traditional barriers to entry.
Advice for New and Aspiring Subnet Owners
- Gareth offers advice to those considering launching a subnet on Bittensor: meticulously plan every facet of the project. This includes defining precisely how the subnet will function, designing the incentive mechanism, strategizing miner utilization, developing robust tokenomics, identifying the target market, and formulating a clear go-to-market strategy.
- “Planning is really, really important... end to end planning from concept to end user.”
- In the current Bittensor landscape (which he refers to as the “detail world,” likely meaning the post-DTA or more mature phase), demonstrating clear and tangible value to token investors is absolutely critical.
- With a growing number of subnets launching, a compelling core idea, a well-defined problem-solution fit, and effective, clear messaging are essential to stand out and avoid being overlooked.
- Actionable Insight: Aspiring subnet creators must present a comprehensive, well-researched, and thoughtfully constructed plan that clearly articulates the value proposition to both end-users and potential investors to achieve success in the increasingly competitive Bittensor ecosystem.
Reflective and Strategic Conclusion
- This episode reveals Vidaio's pragmatic strategy to dominate AI-driven video processing by leveraging Bittensor's rapid innovation for superior compression and upscaling. Crypto AI investors should monitor Vidaio's enterprise partnerships and benchmark performance, as its cost-saving technology could significantly disrupt traditional video markets.
