Department: decentralized physical infrastructure networks
Although the backdrop, in theory, try to provide real usefulness to crypto, there are few who really solve real problems, have a wise commercial model capable of disturbing existing businesses and cannot be easily usurped. Most are simply solutions looking for a problem. A notable exception is a flight tracking network called Wingbits. For what? Because it solves a web2 problem by solving it with web3 incentives. For all those who have already followed a flight such as BA117 from London to New York, you may have used websites like Flightaware or Flightradar.
Figure 1: Wingbit flight tracking card
Source: Wingbits – Transform flight monitoring.
Flight monitoring companies generate millions of income by selling theft data to aviation companies and buyers like financial analysts who monitor private jet movements for mergers and acquisitions. These companies also obtain income from advertisements and subscriptions on their platforms. However, their capital expenditure does not include expenditure on infrastructure and important equipment. Indeed, aeronautical surveillance technology, called ADS-B receptors, is equipment that requires antennas and raspberry IPs, purchased and configured by aviation enthusiasts. These enthusiasts are not long in return, often receiving a free subscription to their favorite flight tracking platform.
The main problem is that enthusiasts are not encouraged to maximize the quality of the data on these networks. Without marginal incentives, ADS -B receptors are often poorly placed – for example, in the corners of the salons or exasperated in densely populated urban areas, leading to low coverage in rural regions.
Figure 2: (LHS) Traditional ADS-B receiver, (RHS) Miner wings
Source: Wingbits – Transform flight monitoring.
Wingbits is revolutionizing flight monitoring by encouraging amateurs to set up stations strategically, based on altitude, while using a system similar to the hexagonal hierarchical index of Uber. This approach guarantees optimized coverage, better quality data and, above all, fair rewards for contributors to the network. They obtained a coverage of 75% of the largest networks with only 1/11 the number of horns. This high level of efficiency, combined with an expected deployment of more than 4,000 stations, should exceed traditional flight monitoring networks by significant margin, providing better quality data to end customers.
The next conversation for family dinner explaining this concept will come easily because we can now indicate a case of real world use, motivated by cryptographic incentives, that everyday people can understand.
Crypto x ai
Similar to market cycles, demand for pic and hollow calculation experiences. GPUs can be expensive and food constraints make them even more. Unlocking the inactive calculation on the general public devices is not a new concept, but the resolution of the synchronization challenge on several devices is. Exo Labs is a pioneering project that carries out breakthroughs in Edge Computing, allowing users to execute models on consumer quality devices, such as household macbooks. This means that sensitive data remain under your control, reducing the risks associated with storage or processing based on the cloud.
Figure 3: A 9 -layer model is divided into 3 bursts, each operating on a separate device
Source: Transparent benchmarks – 12 days of exo, exo labs.
Exo Labs has developed a new software infrastructure called Pipeline Parallel Inference, which allows a large (LLM) language model to be divided into “fragments”, allowing different devices to execute distinct parts of the model while remaining connected to the same network. This approach offers various advantages such as reduced latency, improved safety, profitability and, above all, the benefits of confidentiality.
The exploration of confidentiality also reveals Bagel AI, a project that has developed Zklora (adaptation with a low level of zero knowledge), an approach preserving the confidentiality of LLM with fine adjustment. This innovation allows the creation of specialized models for industries such as legal services, health care and finances, allowing sensitive data to be used for learning to strengthen without risking confidential information.
Although preserving confidentiality is a burning subject, a more important challenge for most LLM is the problem of hallucination, a response generated by AI which contains false or deceptive information presented as a fact. A portfolio manager told me one day: “Wisdom lies in the synthesis of competing points of view to discover the nuanced truth between two extremes.” BlockSense is a project that has developed a owner approach called Zkschellingcoin Consensus. This method aims to superimpose subjective truths of several sources – say, different LLM – to reach a single common truth. For example, imagine execute the same request on Chatgpt, Claude, Grok and Llama. If a model provides an incorrect output, it is statistically unlikely that the four models generate the same false result compared to each other.
Figure 4: Overview of the ZKSCHELLINGCOIN consensus
Source: BlockSense Network – The Rollup ZK for programmable oracles.
Zkschellingcoin consensus could also be applied to the addition of verifiability to AI inference. For example, how can we confirm that an AI agent has properly punctuated USDC in the highest safe at the time of execution? Confidence in AI would be considerably strengthened with an additional verification layer. If we can solve this problem without compromising the cost or latency, this could lead to a major breakthrough in the use of the real world.
The journey from media threshing to reality in Depin and AI shows that a real innovation lies in solving real world problems with practical and effective solutions. Projects like Wingbits and Exo Labs prove how blockchain and AI can create a significant impact – whether by revolutionizing flight monitoring with strategic incentives or unlocking the power of general public devices for a secure and profitable calculation. With advances like Zklora for confidentiality preserving AI and Zkschellingcoin for a verifiable truth, these emerging technologies are about to meet critical challenges, paving the way to a more decentralized, effective and verified future.
