Enterprise hits and misses - AI-for-good versus AI-for-yuck, edge computing hype versus reality

Lead story - applying AI across industries - use case lessons

MyPOV: The overriding view on practical AI is that your data set is a big limitation. So I was interested to check Madeline's Porsche Formula E team races to success with AI, billed as "How to build a machine-learning system with limited data and a mixed bag of talent."

Yes, you have my attention with that one. This use case is still a work-in-progress, but the challenges of bolstering real world data with realistic training data, and earning user trust, are worth airing. 

Let's shift to retail with Stuart's Drinking from the AI fountain at H&M - democratizing AI in retail. H&M's prior AI approach wasn't really scalable. How did they change that? Stuart's quotes H&M's Head of AI Foundation:

Everything we do rests on data...data enablement is the key for delivering on these values that we're looking at. So it's around having proper Master Data Management, having a federated data model that enables all the product teams to work autonomously on the data catalog, data lineage, compliance etc, making sure that those are in  place. But when we talk about the Fountainhead, we talk about the capabilities…When you're doing ML ops, you shouldn't just start from scratch with something; you need to have building blocks in place. That's what we have done.

Next up: AI in financial services. Stuart again: AI in financial services - what sort of future do we want to see? Some considerations from The Alan Turing Institute. AI is already well into real world use in many industries, financial services included. And yet, as the Turing Institute warns, the potential for misuse and harm is absolutely real. We've rolled out powerful tech, with impact on constituents we don't always grapple with. Stuart quotes from the Turing Institute report:

If firms are deploying AI and machine learning, they need to ensure they have a solid understanding of the technology and the governance around it.

That may border on the obvious, but each week, we see evidence that the obvious must not be so obvious: Facial Recognition Failures Are Locking People Out of Unemployment Systems.

Diginomica picks - my top stories on diginomica this week

• A case for cloud repatriation, but let’s be careful before extrapolating to mainstream enterprises - Kurt parses a provocative paper on taking your cloud services back in house, assessing its enterprise relevance.
• Supporting trans people and victims of abuse through intelligent web services - Cath pulls three different NGO use cases into a stellar piece: "The judicious use of technology, applied appropriately, can help make a significant difference to the lives of thousands of vulnerable individuals."
• Retail in the Vaccine Economy - We're still learning what consumer behaviors are going to stick. But, as Stuart reports, poor customer service is not going to fly: Consumer patience with blaming COVID for bad service just ran out - one bit of the 'old normal' makes a comeback. Stuart riffs on data-driven consumer insights from Kroger in Retail in transition - Kroger's omni-channel clout pays off as shopping patterns take shape for the Vaccine Economy, including stats I wouldn't have expected.

Vendor analysis, diginomica style. Here's my top choices from our vendor coverage:

• Can Cisco reinvent Webex to become the verb for hybrid working? - Don't tell Cisco that Webex is legacy; they have other ideas. Phil: "Compared to when I last spoke to Patel in a Webex call a year ago, I found the user experience massively improved."
• Oracle exits fiscal 2021 on a high as Larry Ellison goes SAP-baiting one more time - Stuart on a company that always seems to makes its earnings calls interesting.
• Do we really need a data lakehouse? Hashing AI, cloud, and customer proof points with Databricks CEO Ali Ghodsi - Wherein I liken the term "data lakehouse" to a "mashup of bad Sandra Bullock movies." And the discussion proceeds from there...
• Overcoming innovation challenges in manufacturing with Google Cloud and Siemens - Derek continues our coverage in the run-up to the Google Cloud Digital Manufacturer Summit.

Sapphire Now 2021 is a wrap, but if you're still in a SAPpy mood, we'll have more for you as ASUGForward kicks off next week:

• Sapphire Now 21 - rethinking life insurance for the digital age with Haven Life - Derek
• Sapphire Now 21 - practical tips on Vaccine Economy talent management success from Tenneco and Qualtrics - Madeline

Jon's grab bag - "Employee experience" is the cure-all-du-jour, but how do companies finally move that needle? Madeline shares tips (and data realities) in HR and consumer tech - applauding best practice in employee experience. Stuart asks the kicker question in EU, US tech relations get a photo opp boost, but will this build the digital society we want or just the one we deserve?

Finally, for those weary of buzzword flatulence exuberant pitches about how the edge will solve everything cloud didn't solve, Derek issues a practical view in Throwing cold water on the edge computing hype: "Wanting to skip over cloud and head straight for edge is near impossible, and buyers should instead be focusing on their outcomes and customer needs."

Best of the enterprise web

My top seven

• Low-code and no-code ready for sophisticated application development? Low-code marketers might not dig this one, but Joe McKendrick asks the right question. Short answer: you're not getting away from your IT department that easily, even if you're not a change order nightmare from LOB on their help desk every day.
• RISING with SAP – 4 Key Considerations Before Transitioning to RISE - No one has done a better job of blogging on the brass tacks of effective SAP RISE negotiations than Upper Edge.

Next up, we've got a good AI/bad AI mashup. Let's start with AI-for-good:

• Google Hopes AI Can Turn Search Into a Conversation - "Google’s vision for search that goes beyond a list of links and could change how billions of people search the web." Interacting with Google Assistant isn't much more useful than striking up a conversation with "Marty" exactly intuitive, so there's work to do here, but those businesses reliant on search should be paying attention.
• Artificial Intelligence for Rapid Exclusion of COVID-19 Infection - "If validated prospectively using smartphone electrodes, this will make it even simpler to diagnose COVID infection." That would sure beat the Total Recall nasal swab test I experienced recently.

AI-for-yuck:

• AI still sucks at moderating hate speech
• Facial Recognition Failures Are Locking People Out of Unemployment Systems

It's not a setback that AI isn't good at certain sophisticated tasks yet. But: it's no comfort knowing both of these use cases are widely deployed. Meanwhile, the collateral damage of people wronged by immature "AI" piles up. Enterprises would be wise to avoid this kind of overreach.

Project Health Check is Key to Business Transformation Success - It's astonishing how few projects employ the kinds of independent health checks Eric Kimberling of Third Stage advocates here.

Whiffs

At this point, making fun of self-driving cars is almost too easy to qualify as a whiff. But if Kurt's gonna do it:

I keep forgetting about this keeper via our own Alex Lee:

Hey, managers need sleep too!  As for the glorious future of hybrid work:

Might be time for "no Airbnb Wednesdays," and add Mon, Tue, Thur and Fri to the list while you're at it. Finally, this isn't really a whiff, nor am I sure of the story's veracity, but it's a nice upbeat headline to leave on: Dog ejected from vehicle in Idaho crash found herding sheep. See you next time...

If you find an #ensw piece that qualifies for hits and misses - in a good or bad way - let me know in the comments as Clive (almost) always does. Most Enterprise hits and misses articles are selected from my curated @jonerpnewsfeed. 'myPOV' is borrowed with reluctant permission from the ubiquitous Ray Wang.