Lemon Baird first published his work on Hashgraph’s consensus in 2016, positioning him as an alternative to traditional blockchain architectures. With an computer experience and a career that covers both the academy and the industry, Baird co -founded Hedera to market technology.
His academic career is remarkable for his early work in neuronal networks and reinforcement learning during the 1990s, a period in which the investigation of ia navigated what would later be called the “winter of AI”. Since then, the Hedera project has evolved in a landscape full of distributed competitors distributed approaches, each that claims the technical superiority and addresses different market segments.
Baird, a speaker in consensus 2025, easily transition between technical explanations and commercial strategy, reflecting the dual challenges of building a novel technology and a viable ecosystem around it.
This interview has been condensed and slightly edited for clarity.
COINDESK: His hashgraph algorithm emerged in 2016, during a period in which many alternative consensus mechanisms were proposed. What technical limitations of the previous approaches were trying specifically addressing?
Baird: I love computer science and mathematics side, invent things and solve problems. When I became an entrepreneur 25 years ago, it was the same process. The core of what I always do is try to understand the fundamental problem we are trying to solve. What is the real question? What are we really trying to achieve? And then it is based on that and solves that problem. In blockchain, the fundamental question I asked was: Bitcoin is great, but it is slow and not as sure as it could be with ABFT [Asynchronous Byzantine Fault Tolerance]. It burns a lot of energy and is not as flexible as we might love. I wondered if, in the lower layer in the consensus itself, there could be a way to avoid burn a lot of energy while it is still fast and safe. Is it possible to achieve ABFT, the strongest type of security, while it is also super fast and does not burn electricity or throwing carbon into the atmosphere?
I started working on this in 2012 as one of the many mathematical problems I was exploring. Initially, I was convinced that it could not be done. I would pick up the problem, play with him and convince me that it was impossible, more. But in 2015, I realized that by throwing two hashes, suddenly everything falls in its place. It can have the final speed, essentially at the Internet speed, while having maximum security with ABFT. And it is a stake test, so it does not waste electricity.
From a commercial perspective, the question was: what is the correct way to govern this? When we look at blockchains, they often claim: “We will not have any government. Anyone can help do it.” But the power over time can be consolidated: you end up with a handful of developers or people behind the scene who control everything.
With hedera, we start differently. We made governance decentralize from the beginning. We brought some of the world’s largest organizations: universities and ceiling companies extend worldwide that people trust and have reputation to protect. They balance each other, creating controls and balances, and together they govern the system.
It was about addressing the fundamental question: what do you really want in the government? What will give you a real without confidence, or at least a lower trusted bar necessary to fully trust the system? That was our answer: commercial questions fastened with the same rigor as mathematics.
COINDESK: has mentioned the RWA token, carbon credits and stables as key use cases. These are areas in which almost all the main block chains focuses. What specific implementations in Hedera have demonstrated significant transactions or user adoption volumes?
Baird: I would highlight four key areas:
First, AI is extremely exciting at this time. The dangers of AI are also worrisome, so we need to establish origin, governance and control of versions for AIS. People need to know if they can trust what is happening. Hedera help with the AI in multiple ways, including data permit and the potential handling of royalties for people who provide training data. The work that Eqty Lab is doing with Nvidia and Intel in Hedera is particularly exciting.
Secondly, real -world asset tokenization is transforming the way we handle the valuable assets. We have numerous projects tokenizing real estate, gold, diamonds, carbon credits and even carbon emissions in hedera. From the beginning of blockchain technology, I have maintained that the important thing is not images of monkeys or games, it is that all value things on the planet will finally be put in these accounting books.
Third, the stable are essential if you want the adoption of the real world. We have created a stable currency study to facilitate the development of Stablecoin in Hedera. The Hedera council includes many financial institutions that do an impressive job with Stablecoins.
Fourth, immutable data records are almost exclusive to Hedera through our Hedera consensus service. This allows messages to be sent to issues with immutable access and recording controls. Companies like Hyundai and Kia use this for broadcasting on their entire supply chain.
Coendesk’s research: UCL on energy consumption compared several networks, but the methodology is significantly important in these comparisons. Stake test systems generally have similar profiles, and differences are reduced to nodes counting and hardware requirements. Does the hedera approach differ fundamentally from other POS networks, or is the efficiency mainly of the current network configuration?
Baird: We have been reflective about energy consumption from the beginning. Everything, from our algorithms to how the nodes are executed and governed, and the fact that we use the stake test instead of the work test, deposited the bases for low emissions from day one.
This created a virtuous cycle. The first adopters seeking tokenize carbon loans chose the green block chain. Then, people who wanted to token emissions and credits wanted to use the same block chain where everyone else were doing a similar job. This steering wheel effect has made hedera perhaps the most popular block chain in the green technology space.
According to University College London, Hedera has the lowest carbon emissions by transaction of any block chain. We also buy carbon credits to be negative to carbon. Being green was baked in our structure from the beginning, so we have become a leader in this area.
COINDESK: We are seeing many projects that try to combine AI and Blockchain technologies. Given the different computational paradigms in which these systems operate, what realistic integration points do you see beyond the marketing narratives?
Baird: The intersection of AI and Blockchain is more significant than most people believe. In Hedera, we are seeing a real traction in several areas:
Providence and governance are critical. As we enter a world where everything will be generated by AI, we need to know that we can trust AI. This requires digital signatures to verify the origins, whether human or created by AI.
Data permit is another crucial intersection. When thousands of people contribute small amounts of data to train an AI, each person needs control over their data, the ability to give or remove permission.
Looking towards the future, I am very excited about the use of hedera for identity and incorporate identity into AI systems. We are reaching a point where the means generated by the AI of reality cannot distinguish. The only solution is digital signatures: the content must be signed by the photographer or the reporter. But then you need trusted identity systems to verify those firms.
COINDESK: Having studied neuronal networks in the 1990s, long before the current AI boom, what is your perspective of today’s large language models? Has something fundamentally changed in technology, or are we just seeing the results of the scale?
Baird: Many AI developments have developed exactly as expected. When Alphago defeated the world champion of Go, when Alphazero dominated the chess, when Ais conquered the poker, he had anticipated everything. They even used almost the same techniques they had imagined. We just needed faster computers.
Autonomous cars are also progressing precisely as I predict, using the methods.
But chatgpt and large language models (LLM) have completely amazed me. The 2017 transformer architecture, described in the article “Attention is all you need,” represents an advance that no one could have anticipated. In the 90s, we were completely perplexed by language processing, trying several approaches, but failing at each step.
The capabilities of the LLM today still surprise me, and their future is still unpredictable. Will they reach the superintelligence? Or will they hit a roof? I don’t know, and I would affirm that no one does.
Humanoid robots have also exceeded expectations. While their physical development has matched my predictions, their conversation skills, helpless by LLMS, have far exceeded what I imagined as possible. In the coming years, they will begin with the basic factory work before advancing to trades qualified as welding, plumbing and electrical work.
These technological advances will cause the industrial revolution to seem less compared. Most people have not understood the magnitude of these changes or how fast they are approaching.
