Lead story - we need responsible AI, but does AI have an adoption problem?
I was surprised by the title of Chris' latest, AI’s ‘long tail’ is preventing mature adoption, says Andrew Ng.
I don't see AI as having an adoption problem - Ai has the opposite problem: governance, ethics, and proper design lag far behind AI adoption, generating an exhaust fume of concerning news headlines. But Ng, who was speaking at Bosch ConnectedWorld 2022, knows his stuff. So what's Ng's position? Chris quotes Ng:
I think the biggest potential of AI still lies ahead of us, to use it for all the other industries other than just consumer software and internet. Everything from retail, travel, transportation and logistics, automotive and assembly, and many, many more.
But candidly, when I walk around everywhere from factories to hospitals, they just seek mentors. I think the adoption of AI in all of these industries is still very nascent. A study by McKinsey estimated $13 trillion worth of value annual value by 2030. But I think there's still a lot of work ahead of us to create that value. AI cannot reach its full potential until it's accessible to everyone.
Ng says the most commercially successful AI adoptions are online ads and search, by a wide margin - and they benefit from huge data sets. Ng sees a solution: open source AI tools, powering a "data-centric AI" that will, in his view, "democratize" AI projects. Chris quotes Ng again:
The key thing about this journey to democratize access to AI is to not just let the $100 million or billion-dollar systems get built, but to let all of these $1-5 billion projects in the tail get built.
With data-centric AI, it’s about providing training to more people, but to subject matter experts rather than to machine learning engineers.
Ng is probably right about the need to democratize AI, and make AI development feasible for smaller teams and projects. But we shouldn't minimize the amount of concerning AI adoption already in play: most notably, on social media algorithms that influence elections. But I'd add AI medicine, insurance and HR screening tools. This doesn't necessarily conflict with Ng's views - only with what was quoted from this particular event. Chris concludes, rightly, that it makes sense for companies to avoid recreating the AI wheel, "when what’s needed is more data about the wheel itself."
Intel seems to be in tune with my concerns, via Mark Chillingworth's latest, Intel takes responsibility for AI. Mark writes:
Lama Nachman, Intel Fellow and Director of Intelligent Systems Research Lab at Intel Labs, led the development of the Responsible AI charter and says:
'We see the unbelievable potential of AI. AI is really transforming what we are able to do in drug discovery and climate change, for example. However, if you don’t address the risks of AI, then you are not enabling the potential.'
Intel's "responsible AI" push posits that what is threatening about AI is not Hollywood-style Terminator storylines, but comparatively "banal" scenarios, similar to the ones I listed above (though if "AI" rejects you from a housing loan, it's certainly not a banal experience for you):
These science fiction scenarios can distract us from the very real but more banal ways in which poorly designed AI systems can harm people.
Indeed. On the other hand, the pitfalls of AI ethics are plenty - and platitudes are easy. I'll get back to that...
Diginomica picks - my top stories on diginomica this week
- Never waste a good recession - Uber and Lyft as exemplars of the Gig Economy in the Vaccine Economy - and Stuart never wastes a good headline... "[Uber's] argument about the business model benefiting drivers in a recessionary environment is an interesting one, although one that’s bound to fuel pushback from critics of the company."
- Go to market smarter - G2 makes a bold pitch to vendors - I have mixed views about the mega-software-review sites - I get more honest views on vendors from Glassdoor - but I do think that, on balance, they help software buyers, especially with less complex software decisions. Market intelligence is an interesting G2 move - Barb's on the case.
Vendor analysis, diginomica style. Vendors aren't done with us yet, as fall events roll on:
- COP27 - diving into Salesforce's Blue Carbon framework - Madeline on Salesforce's ESG push, announced this week at the COP27 climate conference. Also see: Madeline's COP27 - why Salesforce is among tech giants hitting the accelerator on climate change initiatives.
- Celonis moves beyond process mining X-Rays towards process mining MRIs - seeking out interdependencies - Derek kicked tires on Celonis' news items: "I think the thing to really watch out for, if Celonis can make it work, is how it could become a platform play for other companies to spring up around its process mining and execution capabilities."
- Sage and Sage Intacct at the crossroads - why Sage Transform marked a change in cloud direction - A frank talk with Sage CEO Steve Hare was one factor in my rethink of Sage's cloud migration strategy, and the future of modern ERP.
A few more vendor picks, without the quotables:
- Zoomtopia ‘22 - Zoom makes a big push beyond meetings into a wider world of enterprise collaboration - Alex
- New Relic on the rise as consumption model transition continues to pay off - Stuart
- GitHub Universe - 94 million happy coders and success for open source, says CEO - Chris
- VMware Explore - moving from "cloud chaos" to "cloud smart" to speed up "the great re-platforming" - Mark Samuels. Also see Mark's VMware use case, How SNC-Lavalin uses VMware’s virtual desktop infrastructure platform to boost productivity.
Jon's grab bag - Barb examined the changing world of the so-called B2B influencer in Influencer Marketing - first, catch your influencer... I'm encouraged by the shift away from follower count obsessions, in favor of customer influence, but I think this conversation needs more emphasis on transparency and disclosure, something that has come, albeit reluctantly, into B2C influencer practices. Meanwhile, if you'll recall that AI ethics teaser, Neil dismantled the AI ethics platitudes in It's time for AI ethics to get real.
Best of the enterprise web
My top seven
- At least $1 billion of client funds missing at failed crypto firm FTX – sources - crypto stole the tech headlines this week, but for all the wrong reasons. On the bright side, maybe this will compel Web3 krishnas to rethink their attempts to tie the future of the Internet to supposedly "immutable" blockchains.
- Phishing-Resistant MFA Does Not Mean Un-Phishable - very useful detail on why MFA isn't immutable either.
- Can You ‘Bot Proof’ Your Applications and APIs? - I see this as more of a daily practice than a "bot proof" end state, but bots are hugely annoying, at times dangerous, and do need a vigorous/informed response.
- CFOs…You Asked For This. Are You Ready? - Constellation's Liz Miller has some strong words for CMOs - and CFOs who want to work with them.
- The Saga of Supply Chain Innovation – Lora Cecere continues an epic blogging year: "While supply chain leaders give lip service building networks, nothing evolved. Companies speak of demand-driven efforts, but market data sits in marketing and sales teams and is not used in supply chain decisions."
- 5 Vendor Sales Negotiation Tactics and How to Counter Them - UpperEdge's Len Riley dispenses his top vendor negotiation tips: "Far too many organizations truly do not understand the sophistication of the executive messaging, sales tactics, and strategies of the vendor community."
- Meta Has Cut More Than 11,000 Workers In Its First-Ever Mass Layoffs - another shoe dropped. I had some spicy arguments with a PR firm over this one. Here's an excerpt from my tart email: "Facebook, meanwhile, overspent on the metaverse, a multi-billion dollar judgement call which must now be paid for, and factors into the headcount reduction. This isn't just some bad weather Facebook is reacting to, Meta helped to create the bad weather they must now cope with." Sharp context can also be found here: Social Media Is Dead.
Twitter's already-infamous blue-checkmarks-for-sale plan garnered loads of well-earned satire. I'm partial to this one:
— rawliberal 🔸 (@rawliberal) November 11, 2022
Another week, another goofy AI writing tool getting overhyped:
Google’s new prototype AI tool does the writing for you https://t.co/inR10UxgYL
I think you meant
"it's unlikely to replace your favorite authors anytime s̵o̵o̵n̵ in your lifetime"
— Jon Reed (@jonerp) November 14, 2022
Oh, and as for my barbs about AI being overly-adopted without proper design: via Clive Boulton, KFC is next up on the sorry-about-our-culturally-insensitive-AI-in-production apology tour. 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.