Getting started with AI and web3: focus first on your partners

But recently, AI has been put in the clutches of the public with the launches of consumer-facing AI products like ChatGPT and Midjourney, and the pending releases of AI browser and search solutions from Microsoft and Google.

AI is increasingly being used together with web3 technologies to create more advanced decentralized apps (DApps) and blockchain-based services. It’s being used for predictive analysis, fraud detection, automated decision making and more: a complimentary technology, evolving alongside web3 and other tech.

Fools rush in

You shouldn’t try to implement any of these technologies for the sake of it. These are new tools for your tech arsenal (and channels for your marketing mix) that can be deployed when you have a good reason. Just as you shouldn’t send an email unless you have something to say, you shouldn’t ‘build a metaverse’ unless there’s a clear value proposition and a solid reason for users to visit.

The implementation of these technologies should be driven by a clear strategy that articulates the vision and objectives you are seeking to meet, how this technology will help you achieve it, the anticipated ROI you expect it to generate, and how you will report on it. These capabilities are enablers, not propositions in their own right.

Start by preparing yourself for this new generation of technologies. Web3 and AI can radically improve your customer experience and business performance, but they are not golden bullets. You can't just point them at your business and, as if by magic, become better at serving customers, selling products, achieving efficiency, or becoming profitable. While these are all possible with a solid implementation of these technologies, there’s some groundwork to do first.

With web3, it’s important to recognize the difference between a campaign and a strategic play and to invest accordingly. Successful, long-term web3 projects depend on the right partnerships. There is a multitude of blockchains, cryptocurrencies, metaverse platforms and cloud computation services to choose from, each with its own merits depending on your goals. Some are more reliable and secure than others; it’s important to seek expert advice from a web3 solution architect to get the right foundations in place.

In the mix

On a tactical level, there is plenty of lower-investment, lower-risk ways to use web3 technologies within your marketing mix for campaigns or advertising, where less consideration is needed regarding long-term technology infrastructure. Creating collectible NFTs as giveaways to grow your marketable database, or delivering an online event using an out-of-the-box metaverse solution, or partnering with the latest web3 video game, are easy entry points to the space, requiring minimal knowledge or effort.

When exploring AI, it’s vital to get your foundations in place to become ‘AI-ready’.

Your content needs to be modeled and tagged for consumption by AI engines and made available to key APIs. This is achievable using either a headless CMS or a modular DXP with a robust content API.

AI technologies use data to learn from and train their models; getting your data in a usable format will lead to better responses and predictions. Good data starts with a data strategy that sets out data sources, data management processes, and how it will be governed in a compliant manner. A data model defines how the data is organized, structured and made available to other systems (like AI engines).

Avoiding incidents: the value of a solution architect

Technology is the enabler for AI, so it’s critical to have the right stack in place. Typically, this includes API-first content management systems, digital asset management, customer data platforms, multichannel integrations, a marketing hub and an appropriate AI engine, plus a cloud platform to provide data storage and high-performance computation.

Commissioning a solution architect familiar with these products to conduct an audit and gap analysis on your current stack and identify any barriers to getting started is a sensible next step.

An area of caution for marketers is the reputational risk that can be brought on by poorly executed AI-driven experiences. Before embarking on AI, it is critical that you consider social responsibility and agree with your ethical stance on AI. Before going live, be sure to check for bias in your model. Data fed to AI engines must be robust and high-quality – otherwise, you can engineer biased outcomes which can be damaging to your brand.

For example, back in 2018 Amazon’s recruiting system assigned ratings to applicants but was skewed by bias against women – because the data it was being fed was historical, mostly from male applications. Similarly, Microsoft launched its AI chatbot Tay in 2016 and had to shut it down after just 16 hours when the bot began to post inflammatory and offensive tweets. According to Microsoft, this was caused by trolls who “attacked” the service as the bot made replies based on its interactions with people on Twitter; either way, it highlights the risks of poorly executed AI. There are hundreds more examples documented on the AI Incident Database.

So, if you get your data, tech and processes in place then you are well-positioned to capitalize on all these exciting new technologies. The world will be your oyster. But get it wrong and you can end up on the wrong side of the press very quickly.

For more hot takes and cold hard looks at the emerging tech landscape, check out The Drum’s deep dive on AI to web3.