Enterprise hits and misses - generative AI wants to take your job, but will it? Edge computing has traction, and (almost) nobody likes Meta's Threads

Lead story - generative AI is a board-level concern - and the jobs debate rages on

Another week, another generative AI jobs debate - along with fresh data on what we know so far. As Derek confirms in Capgemini Research Institute - generative AI tops boardroom agendas, but sustainability is a concern, generative AI is not a vendor marketing confection - it's a board level concern:

Every single board meeting we’ve had this year has had a standing agenda item of AI and ChatGPT. We want to make sure we’ve got our board with us as we’re thinking about where we’re going.

But while most employees will surely be using corporate-sanctioned generative AI tools within a year or two (as opposed to Shadow AI), how much disruption will this cause from an overall jobs perspective? To answer this, we need:

• A deeper and more precise technical understanding of large language models. Yes, we must all become tech geeks to understand the power - but also the limitations - of this technology, which is not all of AI, but one heavily-monetized outgrowth of "deep learning."
• We must factor in the crucial difference between "generative AI does these tasks for me," versus "generative AI can do my (entire) job."

On the second point, Cath's well-considered Generative AI and your job - where will you end up? hits on the nuances of a human job, via a study by Stanford and France's emylon business school:

The premise is that machines are unable to replicate the value of "relational expertise" - or expertise that is generated from, and applied when, working with others in an organisational context - as it is not "extractable or codifiable". As a result, professional roles are more likely to change rather than be destroyed, the researchers say.

Will AI and automation end up taking enough jobs to cause civil unrest, or will it actually help employers fill in talent gaps that inhibit growth? In the short term (1-3 years), I lean strongly towards the latter - but, in a race to the bottom to cost-efficiency, some companies will certainly try to impose this tech as a headcount reduction exercise. Will they succeed? Via the Capgemini study, Derek raises a data point that has been bothering me:

69 per cent believe generative AI algorithms will provide concepts and initial designs, where employees may shift from traditional ideation and creation to review and refinement.

For visual work, yes. But: "review and refinement" can be taken as a less important aspect, when in fact the human-in-the-loop required to remove errors and cross-check generative AI is not just a "refinement" but a risk management necessity, far more important than the initial design itself. And, as someone who has supervised "creative output" since I was a post-college smarty pants, I'm here to say that editing and revision is sometimes 50% of the job, and, often, the most important (human) part of the job.

Derek raises an overlooked point on sustainability:

The concerns around sustainability should be the real eye opener here - I have heard little conversation about how anyone plans to tackle this in any meaningful way.

Derek is right: generative AI energy ethics are neglected. The costs won't be overloooked, however, when executives are shown the cost estimates to operate LLMs at scale. I believe such costs will go down over time - some talented folks are working on the energy consumption for LLM training problem  - but we can add this to the list of adoption/readiness questions.

diginomica picks - my top stories on diginomica this week

• G2 Software Buyer Behavior Study - Informed Buyers want AI and quick ROI, which might mean more Shadow IT... - Yeah, one more AI angle, this one from the buyer side. Barb also notes the declining influence of the software vendor's own sales teams, versus other forms of influence: "Sales are still the least influential source (one percent), down from three percent the prior year. The most influential included industry experts, colleagues or professional networks, online reviews, and internal influencers."
• The topology for team flow - We haven't published many articles on "team topology," but based on this one by Mark Chillingworth, maybe we should: "Skelton and Pais wrote Team Topologies in recognition that technology teams could not meet the demands of the business because they were too disconnected from the value stream - what the customer expects of the organization - whether this be banking, online retail, healthcare or public sector services."

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

• Mulesoft CEO Brent Hayward - orchestrating enablement of the business versus narrowing the developer bottleneck - Martin probes the convergence of tech that might ease the developer crunch: "Modifying a business process to meet new needs is actually best done by the people who understand and work with the process."
• Samsara’s AI lead talks risks, use cases, and explains why LLMs aren’t a silver bullet - Derek shares some thoughtful AI talk via Samsara: "Our role there is to educate and inform and then really try to position our use of technologies, like generative AI, or any of the other sort of new cutting edge technologies, in a way that really solves specific problems for the customers." Indeed.
• Salesforce EVP of Product Patrick Stokes on intentional enterprise friction, 'prompt engineering' and the rise of the Chief AI Officer - Meanwhile, Stuart delves into Salesforce's AI views: "What I mean by that is that nothing that's coming from a generative AI right now is being automatically applied into a workflow. Everything for the time being will require a human in the loop. Everything that we generate and put out, we're going to ask a human being on the other side of an interface to look at it and to say, 'Yes, that looks right to me' or 'No, that looks wrong to me' and move forward from there.

Top customer use cases of the week:

• Spring event highlight - how Esquire Bank makes content personal with Salesforce, and avoids the hard sell blues - Jon
• Morgan Motor Company trades up to IFS Cloud - Phil
• The Folio Society turns over a new leaf with NetSuite - Phil
• Green Cargo CIO avoids being shunted into legacy sidings - Mark Chillingworth

For more context on IBM's future direction - including the Apptio acquisition - check Chris' You don't get fired from buying from IBM, but do you really understand what it stands for? The firm's UK CEO provides some insight.

Jon's grab bag - Stephanie shared the story of a worthy non-profit in Better data is MarViva's hook for more sustainable fishing in the Eastern Tropical Pacific. George wrote about a surprising-but-notable project in Holy Writ - the Vatican and Silicon Valley collaborate on applied ethics. Derek raised issues on the limits of autonomy in Experts warn that absence of human error in autonomous war weapons “is a myth."

Best of the enterprise web

My top seven

• Ahead of AI, this other technology wave is sweeping in fast - Joe McKendrick reminds us that AI is not the only customer priority - managing edge devices is right up there.
• Realizing the potential of live commerce - One big takeaway from McKinsey's live commerce review: different regions have very different live commerce priorities (and maturity levels).
• Why healthcare in the cloud must move to zero trust cybersecurity - Over on VentureBeat, Louis "zero trust" Columbus is still doing his thing.
• What do you know today that you wish you knew when you started your career ? Nothing ! - Vijay Vijayasankar nabs blog-title-of-the-week with ease. Agree on the luck factor, and then we get the quote of the week as well: "Sucking up is not a sustainable career strategy."
• ChatGPT maker OpenAI faces new class action lawsuit over data privacy - The body of legal work needed for companies adopting ChatGPT is far from sorted.
• Questions for boards to ask about generative AI - "Are we balancing value creation with adequate risk management?"
• How Artificial Intelligence is Fundamentally Transforming ERP Systems (and Vice Versa) - Eric Kimberling: "One significant way artificial intelligence (AI) is reshaping our perspective on ERP software is by emphasizing the value of data." Yes, though the customer need for better analytics was the initial impetus here.  By the way, last week Kimberling put me on the video hot seat to ask if I've learned anything from ten years of diginomica...

Whiffs

I'm not sure if Zuckerberg's Threads is going to fail; I suspect it will ultimately be a modest success that will further fragment the social milieu. But it sure has generated some entertaining takedowns, such as this newsletter demolition by Paris Marx:

Threads isn’t even a great app: the experience is clunky, search doesn’t work well, notifications are an absolute mess, there are no direct messages, it’s hard to find your Twitter friends (who are often very different from your Instagram friends), and the only feed available to people is an algorithmically-generated cesspool filled with brands, grifters, and influencers you’d probably like to continue forgetting exist. You can’t even delete your account without deleting your entire Instagram profile

Elsewhere in the universe:

Bonus points for a sensational graphic. Finally, far be it from me to knock Apple's device track record, but I'm going to knock Apple:

Melissa Swift took the design problem down to its core:

True. But it's hard to resist this snark: no one should be excluded from annoying their friends, roommates, and family members. 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.