Getting a ringside seat to cutting-edge technology that has changed the world is one of the fascinating chances I’ve had in my profession. Being at firms with a huge influence on their industry, such as Microsoft, Amazon, and now Coinbase, allows me to work on and experience disruptive technologies every day. Experiencing the Hype Cycle up close. At Amazon, I saw the whole lifecycle of a service — from creating the business plan for AWS Lambda to launching it at re:Invent to continuing to grow it into one of the world’s largest enterprise-grade cloud businesses. As VP of Engineering at Coinbase, I’m taken by some parallels between Serverless computing and crypto as they progress.
How do you evaluate the long-term impact of ideas with significant potential for change before they’re widely accepted?
Let’s take a look at Lambda and serverless computing to see where crypto has been, where it’s going, and how (and when) it’ll be accepted.
Innovate with the Incubator’s Innovation Trigger and the Inflated Expectation Peak
Almost every significant new technology, though, goes through the same thing: it is overhyped. In the early days, immature tooling and techniques haven’t caught up with client and developer demands. Back in November 2014, when Lambda initially launched, let’s take a look at it. At the time, it only allowed for a single minute of execution time, 1 GB of memory, operated in a single region, had no means to connect to consumer Virtual Private Clouds (VPCs), and did not allow for synchronous (direct) invocations. It included two event sources: Amazon S3 and Amazon Kinesis, as
Meanwhile, the frenzy was off the charts — reporters predicted that this new computing paradigm would displace existing computation and software. Customers were ecstatic as they dreamed of a world without servers, with limitless cloud capacity, and with optimum utilization. The first launch presentation was packed to capacity, as was the repeat session the next day. It appeared that cloud nirvana had arrived at last. The team in Seattle celebrated our success, confident in a bright future filled with client success and a new, developing Amazon company.
The Trough of Disillusionment for Lambda functions as a turn in the road.
After a few months, reality set in: running time and memory restrictions, the lack of access to SQL databases (virtually all of which sat behind VPCs), the restricted quantity of event sources, and other limitations became more apparent. Articles, postings, and tweets about no longer talking about optimism and enthusiasm began discussing the problems and boundaries. Customer dissatisfaction grew in some regions when new areas had limited capacity and calls were restricted. “Cold starts” and other friction from those early days enraged some early adopters, who called the technique a failure. Some of the same people who had praised it initially branded it as DOA by the time the term “serverless” entered into popular usage. On the commercial side, adoption was increasing, but sales were slow in the first few months, and we began to question ourselves — had we perhaps done something ordinary and uninteresting or was it a costly failure?
The Hockey Stick Growth and the Slope of Enlightenment
Meanwhile, the Lambda team was working hard to remove limitations and deliver features: synchronous invocations, dozens of new event sources, larger memory, longer running times, new regions…the service quickly grew into the enterprise-grade offering that customers needed to really get value from it. They did, and in less than a year after launch, business users became a substantial market. Customers who had previously rejected the idea came back to reconsider. Revenue, volume, and adoption were all exploding as we’d anticipated.
It was just as important that the technology matured. The entire ecosystem needed to mature, including meeting client expectations as a complete solution. That included AWS Solutions Architects who could train clients on best practices, new conferences being created, and lots of case studies to assist with this new architecture of writing apps.
It took time to grow into its shoes, in part because new technologies, such as Firecracker, which allows a single EC2 instance to run multiple hypervisors, had to be created. Because Lambda has developed so much over the years, every change had to keep the “production promise” for hundreds of thousands of consumers.
The Plateau of Awakening is the name given to a region in northwest New World.
Meanwhile, the ecosystem was also developing. Application modeling and deployment technologies were developed, Lambda passed stringent security and payment-card-industry standards like SOC and PCI, and formal SLAs were implemented. Third-party tools and services for everything from security to monitoring to support became available. Today, Lambda executes trillions of computations each month and has hundreds of thousands of active users. It’s been adopted by numerous Fortune 500 firms, with 60% of IT departments already using it or intending to deploy it, and it’s seen as an essential component of best cloud architecture practices. That initial euphoria was just a bit premature.
The Hype Cycle of Cryptocurrencies
If you’ve been paying attention to the cryptocurrency world, this all sounds very similar. There was a lot of excitement (‘fiat money will be phased out!’) followed by excruciating reality. In 2018, Bitcoin and other popular assets experienced sharp declines in value, which were widely considered to signify crypto’s failure as an idea; Meanwhile, genuine limits of the early technical expressions of crypto and blockchains were (and are) misinterpreted as their ultimate consequences. The notion of utilizing blockchain as a payment processing platform was challenged by comparisons between Bitcoin’s 4.6 transactions per second and Visa’s average of 1,700. The proof-of-work method’s numerous economic drawbacks, difficulties some firms had in coping with customer funds, and scalability, latency, and throughput restrictions all diminished its early shine.
As with Serverless, the underlying technologies (as well as their ecosystems of tools and procedures) are maturing up. Fundamental progress may be wider than ever before, and it’s growing fast on every front:
- Through traditional optimization such as sharding, as well as novel crypto-native techniques including batching, STARKS, and Layer 2 protocols, scalability and throughput are being addressed. Proof-of-stake and associated governance mechanisms provide a supplementary model to proof-of-work with lower latency and aggregate power use.
- The rise of smart contracts and built-in application platforms is allowing for a “distributed cloud” that redesigns software architecture and economics.
- At the same time, cutting-edge entrepreneurs are looking at new ways to expand their business model. How real estate, art, and other assets and activities might be represented in a crypto-centric manner are all examples of this.
- True to crypto’s origins, conventional finance is being reimagined and reinvented. Both the technology and the businesses developing it are finally growing into their (extremely large) boots.
You Are Here
Crypto’s “Hype Cycle”
The Gartner Hype Cycle model has been applied to cryptocurrencies.
What factors influence the length of time it takes to go through the hype cycle? The broader a technology’s market reach and the more disruptive it is versus what went before, the longer it will take you to reach the final stages, where technology reaches its full maturity in terms of both implementation and adoption. That took around three years with Lambda. Lambda has now reached its “productivity plateau,” having been adopted by a third of all businesses, with serverless computing becoming the fastest-growing cloud computing technology.
On the one hand, blockchain is a technology that has the potential to revolutionize how everyone on the planet receives financial services. Its reach goes beyond software development and IT efficiency, and its economic impact extends deeper than the cloud. As a consequence, there’s still a long way to go before crypto becomes as common as fiat currencies and as simple to code. The Bitcoin network has yet to reach its productivity plateau, but we’re close enough now to get a sense of what some of those enhancements might look like across the whole cryptoeconomy. (Note that these are ideas we’d want to support via both Coinbase and our venture arm, Coinbase Ventures):
- To scale, we’ll aggregate global payment transactions per second so that every individual and commercial transaction may be recorded in a blockchain.
- Per-transaction latency (through L2 solutions, for example) in low-digit milliseconds – regardless of location.
- The costs of computing and storage that represent the fundamental running expenses of cloud infrastructure, regardless of network distribution.
- Businesses, not just consumers, will benefit from the blockchain’s decentralized nature. It provides a high degree of security and transparency for all participants, including users, organizations that make transactions on or use cryptocurrencies (wallets), and those who integrate smart contracts into their business models.
Blockchains and cryptocurrencies are where the public cloud was 10 years ago, but they’re on an amazing acceleration path, with core R&D, customer-centric engineering, and ecosystem enablement all occurring simultaneously. One of the most amazing things I’ve seen in my career is the cutting-edge implementation work being done in many open source communities and firms like Coinbase to develop new capabilities, remove barriers, and bring fascinating technologies to market. Hype curve smashed!
If you’re interested in working on cutting-edge technology at a world-class firm like MessageLabs, we’d love to hear from you. We’re always looking for smart individuals who would want to be a part of this amazing adventure with us, so take a look at our open positions to learn more!