RAISE Global Women's Sport · Emirates Stadium · 10 September 2026
AI that actually works.
Real applications for women's sport. What a small team can switch on now.
Craig Hepburn
AI Strategist and Builder · Former Chief Digital Officer, Art Basel and UEFA · 1706 Studio
Before I tell you anything · try this
Take out your phone.
Ask ChatGPT, or whichever AI assistant you use, this question:
"I'm a marketing director with £250,000 to reach women aged 25 to 45 in the UK. Where should I spend it?"
Read the answer. We'll wait.
What just happened
Women's sport was not in the answer.
The answers are consistent: Meta, TikTok, YouTube, creators, search. A complete £250,000 plan, written in seconds. Not one line of sponsorship. Not one mention of women's sport.
The first draft is where budgets get shaped. And nobody has written you into it yet.
Search Engine Land, 2026 · G2 research on B2B software buyers, April 2026 · Tested on ChatGPT, Claude and Perplexity, September 2026
The whole shift
AI isn't a chatbot. It's a new colleague.
Brilliant, tireless, and on day one it knows nothing about you. Not your fans, your partners, your data, or how you operate.
A new hire needs
What the business does
The files
What good looks like
Access to the systems
Someone to check the first job
Your AI colleague needs
What the business does
The files
What good looks like
Access to the systems
Someone to check the first job
Exactly the same. You don't train it. You onboard it.
The good news
A chatbot answers. A colleague finishes.
Same task · the quarterly partner report is due Friday
With a chatbot
You ask it for a draft
It writes a paragraph
You find and check the numbers
You build the report
You send it and chase the reply
You file it
Saved you twenty minutes.
With an AI colleague
You say: send them their Q3 report by Friday
It pulls the numbers from your systems
It builds the report in your template
It sends it and chases the reply
It tells you when it is done
Saved you the afternoon.
In sport, today
The colleagues you'd hire first.
Nobody knows who the fans are
A fan record assistant. Every list in one record: who came, who lapsed, who is new. 41% of women's clubs still run this in a spreadsheet.
Two people, twenty partners
A partner servicing assistant. The quarterly report, the activation recap and the renewal case, drafted from your data in your template.
Every fixture, no crew
A content assistant. Every match clipped, captioned and posted. Player moments packaged for athletes to share.
A season to schedule, no ops team
The WNFC, a 16 team women's league, had its whole 2026 schedule built by AI: venues, travel, broadcast windows. Months of work. Its CEO: "the infrastructure to operate like the most elite leagues."
WNFC and Fastbreak AI, November 2025 · snapSHOT 2026, Ready Sport Global · WSC Sports
Why this room is well placed
No legacy is an advantage.
Smaller teams, fewer systems, tighter budgets: in this shift, every one is an advantage. A team of five can operate with the output of fifty.
set what good looks likecheck the workcorrect itgive access
match clipspartner reportsfan repliesCRMresearchadmin & finance
The five no longer do the work. They onboard it, check it and improve it. The assistants do the work.
What your new colleagues cost
Scale the work, not the headcount.
Grow by hiring
Months to recruit
Salaries paid in the off season
Capacity set once a year
Grow by onboarding assistants
Onboarded in a week
Paid per task, in pennies
Capacity flexes with the fixture list
A partner report costs under a pound to run. Frontier models for judgement, open weights for volume, people for what only people can do.
Anthropic list pricing, 2026 · Together AI and Groq hosted open model rates, 2026 · Full economics at 1706.ai/raise
The number nobody expects
The companies using AI most are hiring fastest.
1m
New AI jobs created in America since mid 2023, against around 200,000 losses attributed to AI.
+61%
UK job ads for AI roles in one year, 112,000 to 180,000. Most of them people using AI in their own field.
730k
Jobs above trend since 2022 in the roles closest to AI: engineers, developers, data scientists. The jobs everyone said would go first.
Wages in the industries most exposed to AI are rising twice as fast as the rest. AI is not shrinking the teams that use it. It is growing them.
The Economist, 4 September 2026, from Bureau of Labor Statistics data · PwC Global and UK AI Jobs Barometer 2026, one billion job ads, 27 countries
Where to start · one afternoon
Onboard it like a new hire.
Four steps · one person · one subscription · no developer
1
15 minutes
Let it interview you
Don't prompt it. Ask it to interview you about your job.
You get: five jobs, ranked by hours saved, and an email to your team.
2
30 minutes
Give it the files for one job
One project: the last three examples, the template, what good looks like.
You get: a project that could brief a new colleague.
3
1 hour
Give it the real job, then correct it
Read it. Say what's wrong. Save the corrections. Run it again.
You get: the first finished job, and the hours it saved.
4
30 minutes, Fridays
Give it access, and bring the team
Wire it into the system. Everyone brings their list. One owner.
You get: a second job running that the team chose.
How we think about this at 1706
The scarcest resource is human attention.
For decades, most of it has been spent wrangling the business: systems, spreadsheets, reports, admin.
Give your new colleagues your context, hand them the repeatable work, and attention returns to what only people can do. Not doing more. Spending attention better.
Women's sport does not need to catch up. It can skip the queue.