By Yemi Olagbaiye, director, client portfolio at digital software consultancy Softwire
“A parent is someone who carries pictures where their money used to be.” Few parents would disagree with this sentiment, especially considering that the cost of raising a child to 18 is now nudging £300,000 or $310,605, according to the latest UK and US figures.
How does this relate to the rise of GenAI? As both a dad and someone experienced in leading AI projects, I can tell you the gap between the two isn’t as wide as you think – particularly on the money pit front.
OK, CTOs probably don’t carry pictures of their AI projects in their wallets, but they might well feel nervous about the cost implications; recent figures suggest the price tag for AI development can be significant, ranging from $50,000 (approx. £40,000) for small to medium projects to $5 million (£3.86m) and over for large-scale initiatives.
Here, I’ve pinpointed five financial similarities between having kids and implementing GenAI, and I’m only half-joking: if businesses don’t consider the very real, often unseen expenditure of exciting new tech, then that AI project they’re cooing over might come with some unwelcome surprises…
The upfront costs aren’t the actual costs
Implementing GenAI and preparing for a baby – especially a first baby – both usually require a large upfront investment, whether signing up for a foundational model or buying prams, cots and baby seats. However, this initial outlay can disguise the ongoing maintenance costs. With babies, that could be the endless supply of nappies and formula, while for GenAI, it’s tool subscriptions, cloud services and model updates.
Further down the line, parents grapple with ever-growing childcare bills and haemorrhage cash on clothes, shoes, and bikes that their children immediately outgrow. Meanwhile, unseen GenAI expenses include having to retrain models over time, making sure they evolve in tandem with the business’s needs, as well as the ongoing costs of maintaining data security, privacy, and compliance.
Beware the opportunity cost
Both parenting and GenAI come with an opportunity cost, requiring sacrifices for future benefit. Once children arrive, most parents give up some degree of personal freedom: it’s generally not easy to spontaneously go out on the town, devote hours to time-consuming hobbies or go on friends’ holidays in the same way.
Implementing GenAI is not dissimilar in this respect: Companies will find there’s a trade-off between short-term disruption and long-term technological gains. AI-driven innovation needs to be balanced with maintaining current business operations and the strategic initiatives it may conflict with. Then there is the time it takes to integrate GenAI with the processes of all the different functions and divisions within the business.
Learning curves cost money
The necessary adjustments and learning curves of GenAI and parenting can both be costly. Parents are all too familiar with the frustration of expensive outlays for new passions, such as musical instruments, club uniforms, sports equipment and classes, only for their child to rapidly lose interest.
How does this relate to AI in business? Well, you can develop a GenAI pilot and allocate a crack team to build this out but suddenly find that results don’t meet expectations or that the business has evolved and its needs have changed. Substantial reworks can be very expensive. Unlike parenting where it’s less possible to anticipate your child’s sudden aversion to trumpet lessons, there are ways in which companies can navigate the AI learning curve effectively. Phased rollouts and iterative testing ensure that GenAI pilots stay connected to the demands of the wider business and that any adjustments can be made in a timely and cost effective way.
Scaling up has a price
There is no denying it: the cost and complexity of GenAI deployment and parenting increase with scale.
The more kids you have, the bigger the car, home or childcare bill. And don’t underestimate the mental load of scheduling: Child A needs to go to karate at the same time as Child B needs collecting from daycare on the other side of town.
Equally, scaling AI across an organisation is rarely linear, whether it’s expanding an AI pilot into various functions or introducing an AI solution across different regional offices. There will also be integration challenges: companies might have more data-processing costs in one location versus another or could need to work on cross-departmental alignment to get everything joined up.
Balancing return on investment with a return on intent
Implementing GenAI and embarking on parenting both involve a leap of faith, coming as they do with escalating and hidden costs, and, of course, a degree of uncertainty regarding ROI. After all, can any parent or CTO accurately predict how their creation will turn out?
While this is clearly tongue-in-cheek, there are definitely parallels. Raising kids can be challenging and expensive, but the intent behind the investment is to raise a happy, fulfilled, and productive adult—and that’s obviously worth it.
When it comes to GenAI, companies need to be clear about their return on intent and return on investment. This involves avoiding only focusing on short-term GenAI productivity wins and instead thinking about aligning AI projects with long-term, sustainable business goals.
