Last night, I had a pre-Valentines Day date with my lovely wife, Lisa Reisman (also my boss and the co-founder of our sister site MetalMiner). Lisa and I both came from the sourcing and supply chain consulting world of yore, and have used spend analysis and related artificial intelligence including auto-classification technology for longer than we’d care to admit (in my case, my work with spend analysis predates our relationship and AI, with manual spend clean-up work in Excel and Access). At one point, I even had the dubious distinction of recommending Ketera to her last firm for spend analysis because it was, well, cheap.
And she got (at least initially) what she paid for, much like my previous short-term relationships. Before I fell for my future wife-to-be, that is. Alas, I survived the beta test before the marriage GA roll-out. And there haven't been any upgrades or migrations since (thank goodness).
Ketera. Let’s just say it was never a lasting affair (for anyone). But even dark clouds have good linings – through Ketera, I came to meet Burton Goldfield, who now runs TriNet, an outstanding Personal Employment Organization (PEO) who our firm outsources HR to, in part because of the initial relationship formed when Burton had a short-term stint at the thing before Kleiner and others ended up selling off the technology and development team to Rearden/Deem.
Anyway, back to the subject at hand -- for our early Valentine’s Day celebration, Lisa and I went to see Her starring Joaquin Phoenix. Spike Jonze directed (of Being John Malkovich and Where the Wild Things Are fame).
Her is probably the most impactful movie of the year in showing what our society is moving toward. It’s profoundly important in signaling a world we’ve already entered and can’t retreat from – a place where applications and networks get smart and learn from our actions as we teach them. And a world where, well, we can even fall in love with our operating system (OS) because it talks back and knows us better than ourselves.
Her is a thing of beauty and irony even without the lovely, raspy voice of Scarlett Johannsson – don't we all secretly want to be her valentine – starring as the uber smart OS, “Samantha.” The protagonist, Theodore Twombly, played by Phoenix, is even a type of artificial intelligence layer himself in his day job – he writes “real” letters for spouses, parents, and kids to other family members and then drops them in the real mail (and like the OS with which he falls in love, he develops long-term relationships as he serves as the surrogate muse). And his BFF, who ends up also falling in love with the same OS, develops video games – also serving as an intelligent layer between life and reality.
I’ll stop right there with the plot and my quick introspective thoughts. Her is amazing. See it. Run to the theatre. I don't want to spoil it for you.
More important from a Spend Matters perspective is what Her signals for our world. Within procurement, machine learning and other predictive techniques are all around us. We just don’t know it yet. If you’ve used any spend classification tool as part of a spend analysis program – or someone has used such a system on your behalf – then you’ve relied on the same class of technology highlighted in the film. When a system learns (e.g., that IBM, International Business Machines, and International Biz. Machines are all one in the same and that Polaris is a heart rate monitor and not a missile system for spend classification purposes) it’s using artificial intelligence at the core.
Next, imagine our overall ERP and integrated source-to-pay procurement systems having an artificial intelligence layer that learns from us and guides us in everything we do. It’s contextual. It’s directive. And it knows how to present the right things at the right time for maximum effect. For example, it senses we’re sitting down just before a flight at the gate and have 5 minutes until boarding. What a perfect time to go through requisition requests on a personal “batch” basis simply by talking into an ear piece with a microphone (which reads us the requests). Or take it a step further, and our procurement-Samantha will flag only the requests that it thinks should not go through on a “straight-through processing” basis.
Or consider another use case. Why not, before an RFI is issued, have our trusty Samantha flag or recommend new suppliers based on patterns in past performance or external market condition data? “Jason, based on the quality issues that are cropping up combined with latent IP theft concerns we’ve had with suppliers in Shandong province, perhaps we look elsewhere for injection molding and assembly suppliers for that new component in the upcoming product launch. I have a few in mind based on moves our competition has made in recent weeks, based on new container data coming from China and US customs reporting.”
Scary. But this is not just “potential” technology. It could be done today, and will most certainly be done tomorrow.
Finally, if you go see Her, don’t miss the following geek notes:
- The retro-kitsch smartphone device with camera. It’s high tech but it’s not. Really, the device continues to be a fashion accessary, much like the iPhone has become, but in a more retro way sort of like the Leica digital camera is today. Curiously, there’s no “Google Glass” type device in the movie (but cameras can be as small as a tiny dot which can be affixed anywhere)
- A lack of keyboards in the movie. Voice and hand gestures rule the day of future interfaces. No surprise!
- How Samantha gets progressively more aggressive in prompting action (vs. just responding) as she learns. This is a scary but natural outgrowth of AI and related machine learning evolutions.
- The importance or perception of the hand-signed note as well as emotion that can only be conveyed through human-authorized prose. Computers aren’t yet smart enough to duplicate our writing, but they can, as a feel-good plot twist shows, curate it for us!
On a final note, just make sure you don’t fall in love with your SAP/Ariba, Oracle, Coupa (or other) system just yet that way. Although we’re sure the future “voice” of these systems will likely be selected in a manner such as the best pre-sales engineers are today – for their overall appeal, not just their ability to fire up a demo with the best of ‘em.
Yes, I’m still a pig, I admit it. Though a loyal one, I must add.
Happy Valentines Day, everyone.