A Simple Guide to Navigating Deep Tech Innovation and Uncertainty

Apple’s decision in 2024 to cancel its autonomous electric-vehicle initiative, “Project Titan”, spelt the end for one of the most expensive and prolonged technological experiments undertaken by a major firm in recent years.  

Despite a decade of development, the project had neither achieved full autonomy nor articulated a convincing pathway to commercial viability. As Francis D Kim and ISB Professor Rajendra Srivastava argue in their study, this failure cannot be explained solely by technological complexity or softening EV demand. It reflects something more fundamental: Apple misread the very nature of the innovation it was attempting, treating a profoundly uncertain, future-dependent endeavour as if it belonged to the stable, incremental domain in which the company usually operates. 

Misinterpretations of this nature, the researchers contend, are widespread. “Deep tech” has become a catch-all phrase that obscures more than it clarifies. To cut through this blur, they have introduced a typology organised around two dimensions: whether a project relies on continuous or discontinuous technological development, and whether its market already exists or must be created.  

The researchers’ categorisation yields four distinct modes of innovation—pivoting, forecasting, projecting, and backcasting—each defined by a different configuration of risk, and each requiring a different strategic stance. This typology serves as a practical way to understand what each project truly demands, helping leaders align expectations with reality well before development begins.