CIOs are having to rethink the long-standing "In, Out, Here, There" model for their companies. That is, which technologies and people stay inside with internal IT, which go out to a vendor, what stays local, and what moves around the globe.
A generation ago, the one-two punch of the Internet + Cloud sent the "Out and There" quadrant booming, with SaaS implemented and managed by global systems integrators. These SaaS solutions focused mostly on corporate SG&A (think sales and customer support with Salesforce, HR with Workday, IT support with ServiceNow, etc.).
AI is forcing the same four questions again, and this time the answers are very different, as the "In and Here" quadrant rapidly takes priority. Why? Because AI is the new means of production and its value sits at the heart of the business, concentrated in the Cost of Goods Sold (the specific ways a company delivers its unique products and services).
The new In is now the most important vector, demanding a significant "remuscling" of internal IT capability. Coca-Cola, Heinz, and McDonald's were famously built on their secret sauces; going forward, every company's secret sauce will be its own unique application of AI to COGS, which means IT needs to treat the new compute / data / model / solution stack like a state secret. That value can't be outsourced, whether to a consulting firm or a frontier model.
The new Out is everything that isn't the secret sauce. The line between core and context is being redrawn, and context is getting ruthlessly repriced: third-party services at fifty cents on the dollar, SG&A budgets newly eligible for the same treatment. Core stays in. Context goes out, with dramatic deflationary impact from AI.
The new Here is proximity. Because AI's value lives so close to the business, it needs the right people embedded in it, those with a deep understanding of industry combined with sophisticated application of AI: hence the rise of the Forward Deployed Engineer (FDE). Post-Covid, it turns out co-locating your best people matters more, not less.
The new There is the open question. As labor arbitrage economics dissipate, offshore locations need to be reimagined as capability and innovation centers, not capacity and cost, but how that plays out could go two ways. Does it become a "good bank, bad bank" split, with legacy outsourcers stuck servicing old "keep-the-lights-on" tech while a new breed claims the AI work? Or can the FDE function itself be delivered from the other side of the globe (in what would be a "rearward deployed engineer" motion)?
The old IOHT wrote the rules of the road for a $3 trillion tech services industry. If you got caught on the wrong side of them, the market punished you. The new IOHT is writing the next set of rules.
