It's a typical Saturday evening, and while most teens might be out with their friends, our digital "teenager" - Artificial Intelligence - is at home getting high.

No judgments, we were all young once.

The problem isn’t AI's desire to trip balls on huge quantities of data; it's that it presents these hallucinations as factual information.

And for those who have no idea what I’m talking about and assume I’m the one on drugs:

A.I. hallucinations are a phenomenon where large language models produce nonsensical or inaccurate outputs due to over-interpretation of their training data.

There are entire guides on how to prevent AI chatbots like ChatGPT and Claude from hallucinating - typically through refined, specific prompts that limit their responses to more narrowed outcomes.

This is why generative AI is still too sketchy for reinsurance.

Eventually artificial intelligence will be able to manage and optimise vast reinsurance portfolios in real-time, predict every shift in the market, and provide insights no human could fathom.

But we’re not there yet.

Consider a minor error, a decimal point misplaced, a misunderstood policy term, or a single misjudged risk assessment.

That could cause untold losses that ripple through entire reinsurance portfolios.

For an AI to be useful, it can't just be "mostly" right; it needs to be precise every single time.

So, what can the reinsurance industry do while AI flies high with unicorns and fractal spreadsheets?

We can exploit the transformative leap of modern tech, which is miles ahead of the painfully archaic systems cedents, brokers, and underwriters grapple with.

And when I say transformative, I’m talking by orders of magnitude - think 17th-century postal service vs. email.

Coming from a B2B SaaS background, I was flabbergasted that an industry as big as reinsurance still works with software from the floppy disk era.

If It Needs to Be More Than 90% Accurate, Don’t Use Generative AI

Reinsurance is ripe for disruption.

With its maze of contracts, dynamic risks, and vast troves of historical data, the industry is calling out for revolution.

And this is evidenced by the significant rise in reinsurtech startups who are delivering on the promise to transform the value chain for:

Cedents, with improved placements and better pricing.

Brokers, enjoying swift, knowledgeable negotiations while free from janitorial data work.

And for underwriters, a universe of high-quality and transparent information, with really useful risk assessment tools at their disposal.

And this is all without generative AI.

While artificial intelligence has made strides in understanding reinsurance, it's nowhere near ready to take over.

Your job’s safe for now, but are you maximising your potential?

Reinsurance teams wade through endless amounts of data, never conveniently formatted, to prepare for renewals.

Errors, inconsistencies, and missing values.

It’s painful.

Once the submission reaches the reinsurance underwriter, their eyes roll in weary frustration.

And as much as they might want to, they can’t blame the cedent or broker - there’s simply too much data for teams to manage effectively without the right tools.

And no, Excel is not the right tool.

Industry giants like Lockton Re, Markel, and Generali recognise this fact; they’re embracing modern platforms that facilitate comprehensive portfolio insights, effortless collaboration, and accelerated go-to-market speed.

Congrats to whoever’s credited for implementing these tools and boosting the bottom line - I hope you’re enjoying the promotion.

Embrace Today, Prepare for Tomorrow

Reinsurance is a dance between data and human intuition.

A clunky one, currently.

Not surprising, robots have about the same rhythm as a drunk uncle at a wedding.

But make no mistake, we are racing towards an endpoint where AI’s capacity to handle vast datasets will render human efforts quaint by comparison.

Assuming ChatGPT doesn’t suddenly become sentient and decide to destroy us all on a whim.

In the meantime, you can choose not to rage against the machine and have it do all the tedious data stuff that makes you wish a Terminator-style Armageddon would arrive before 5pm.

Then again, making AI run relentless, dreary data duties might actually cause the apocalypse.

Destroying humanity would be less work than that of a reinsurance team.

But we digress.

Regardless of where your role sits in the reinsurance value chain (assuming you’re even that cool), data is always a hassle, whether that’s collecting, preparing, analysing, and so on.

Costly, too.

Luckily, you have the means to extract gold from your data mines without relying on generative AI.

Modern platforms handle huge datasets and aren’t constrained by row limits or the irritation of juggling multiple spreadsheets; they simplify data cleansing and uncover insights that were buried deep in overwhelming amounts of complex information.

Embracing modern tech isn't just about being ahead of the curve; it's about survival.

Those who leverage modern reinsurtech platforms stand to gain lasting relevance and profitability from the oncoming tsunami of AI advancement.

The ones who don’t will drown. Or at least struggle to stay afloat.

While AI matures through its complicated teen years, our industry has a golden opportunity to refine its data processes and position itself for the AI-driven future.

So lace up those dancing shoes and get ready to groove.

Just remember to lead and let AI follow - for now.