Enterprise hits and misses - Microsoft changes their spin on ChatGPT for search, while enterprise uses for Metaverse and generative AI get real
- This week - Microsoft changes their positioning on ChatGPT for search. No surprises here - but the pursuit of enterprise use cases is worthwhile. Plus: time to revisit the enterprise Metaverse. Your whiffs include cheating on AI ethics exams - and data backups for Valentine's day.
Lead story - Metaverse and AI use cases hit retail - will they pan out?
Readers know I'm not a good candidate for Metaverse fanboy of the year - nor do I appreciate the tech media cheerleading unproven AI use cases.
So what is my take on two new retail scenarios? Start with Derek's L'Oréal CEO says the beauty giant is a ‘Unicornus Rex’ set for the Metaverse. What does L'Oréal have in mind? Derek:
On the partnership with Meta, L'Oréal announced that incubator will support at least five startups that specialize in 3D production, augmented reality, virtual reality, mixed reality, avatar creation, portability in user experience, token economy or Rotherham topics related to the metaverse and Web3. Interestingly, it added that these startups don’t specifically have to be related to the beauty industry.
Derek details several Metaverse use cases L'Oréal is pursuing. This one jumped out:
And then L'Oréal’s partnership with Ready Play Me, a cross-game avatar platform for the metaverse, is already in full swing, as two of the Group’s leading brands - Maybelline New York and L'Oréal Professional - have launched makeup and hairstyles for avatar creation that can be used on more than 4,000 platforms and apps worldwide.
I'm a stickler for "Multiverse" over Metaverse, since there won't simply be one 3D community. But I've always felt that fashion was one of the strongest consumer Metaverse use cases - either to virtually try on new looks or, in this case, customize avatars. As Derek notes, L'Oréal is already turning in good corporate numbers, so this type of early-in investment makes sense.
I also see the potential for royalty streams based on paid in-game/in-character beauty enhancements. Where do I part ways with L'Oréal's approach? The bothersome blurring of the lines between Metaverse, "Web3," and NFTs. If L'Oréal gets caught up in heavy NFT market investments tied to crypto, I don't think they'll have super-happy news to share on that later. But the Metaverse play itself makes sense. As for "A Unicorus Rex in the Metaverse," if that's a good rallying cry, I'll step aside.
Meanwhile, Stuart examines a very intriguing AI use case in Why Redbubble has designs on AI to improve the e-commerce content experience for customers. Essentially, Redbubble is using AI not to replace creators, but to make sure the best creations are properly filtered, tagged with data, and surfaced on its platform. Stuart quotes Redbubble:
We are most excited about the opportunity to step-change our understanding of artists' content through the use of AI. This will enable improvement in search and discovery across both marketplaces, both on and off-site, and solve a core, long-standing challenge of how to build an objective, accurate understanding of each piece of content available in the marketplace, so we can place the right design in front of the right consumer every time.
Though this scenario is not yet mature, this use of AI looks promising - sourcing a unique data set and solving the problem of tagging and discovery - at a scale humans simply couldn't. I believe Bing's use of ChatGPT will either be a disaster or an entertaining carnival sideshow, but that's because of the nature of the data set it's been trained, on versus its stated purpose (searching for verifiable/trusted info). However I like enterprise/community-specific search as an AI use case - this example shows why.
While some siren voices will mutter about silver bullets here, this use of AI tech is an interesting real-world application that appears to fit a clear business/operational need. This is an initiative that’s going to be well worth keeping an eye on as the testing/learning phase moves into customer-facing delivery and, it is to be hoped, an uplift in metrics.
That uplift in metrics part cannot be overstated. We seem to be losing track of that criteria with AI lately.
Diginomica picks - my top stories on diginomica this week
- Past the point of keeping the lights on - why data is the real problem when it comes to sustainability issues - Martin's raises a crucial/neglected sustainability question: ""Is it possible or not to actually reduce, halt, or send into decline, the rate at which data is consuming energy?"
- Why marketers can’t afford to ignore the middle of the funnel - Barb issues a wake-up call to marketers: "What happens in the middle of the funnel is the most enlightening area marketers need to focus on because it influences whether someone we want to do business with will move to the next stage."
Vendor analysis, diginomica style. Here's my three top choices from our vendor coverage:
- Salesforce cleans up emerging markets with 280,000MWh renewable energy investment - Madeline on Salesforce's lateset sustainability moves. Also see: Salesforce adds gender identity and pronoun data fields to promote inclusivity in tech.
- Xactly and the new tech world - CEO Chris Cabrera on the macro-economic climate - Brian shares his deep dive with sales performance and revenue forecasting Xactly.
- Are supplier-driven transformations a realistic goal? Putting SAP Business Network's talking points to the use case test - Supplier-driven transformations sound like a worthwhile concept - but is this a viable framework? I dig into two SAP customer use cases for answers.
- Dropbox delivers growth, but records $162 million real estate loss - CEO cautious about the year ahead - Derek quotes Dropbox CEO Drew Houston: "We recognize that as our customers experienced challenges in their businesses, and evaluated their budgets, there's added pressure to reduce software spend. But we see an opportunity to mitigate some of this pressure through more high-touch account management for these customers."
- Pega passes $1 billion in subscription revenue as it delivers strong Q4 - Derek puts Pega's subscription/earning news in context: "However, Trefler didn’t shy away from the fact that market conditions are tough and that there will be macroeconomic headwinds this year. That being said, Trefler argued that Pega is well positioned." Derek adds: "It’s worth adding that process optimization and low-code are front of mind for buyers, as they look for cost cutting opportunities to do more with less."
Enterprise earnings are currently tied to being able to provide a jugular, real world return for customers. However, even in those cases, a sense of belt-tightening pervades.
A few more vendor picks, without the quotables:
- The Tercera playbook for SI success in the next wave of cloud - Phil
- Celonis - ‘We need to look at retail processes as a whole and bridge the silos’ - Derek
Jon's grab bag - Derek rings the warning bell on one of generative AI's most concerning downsides: ChatGPT has the potential to spread misinformation ‘at unprecedented scale’ (going forward, we'll need to distinguish between regurgitated misinformation, intentional disinformation, and/or disputed information with baked-in or unintentional algorithmic bias). Chris turns his attention to the Multiverse, err, Metaverse - in the banking arena: Should we bank on the Metaverse? Chris:
On the face of it, there may be a tension between what some customers want from a metaverse – freedom, anonymity, fun, and play – and what companies would need to provide, in the form of ID, authentication, safety, and security. Put another way: before we enter Wonderland, we will fall down a lot of rabbit holes.
Best of the enterprise web
My top seven
- AI and DevOps, combined, may help unclog developer creativity - Joe McKendrick asks: can modern tooling can help developers do what they do best?
- Tech Was Supposed to Make Cars Safer. It Didn’t Deliver - Though I'm Jonny-downer on fully autonomous driving, I do believe that tech, properly applied, can absolutely make cars safer (assisted parallel parking and pullouts, for one - who couldn't use help with that?). But this scorcher raises the right questions.
- An interview with former Cisco CEO John Chambers - After my prior belt-tightening talk, Chambers' views are a motivational tonic: "There’s a playbook for dealing with an economic downtrend. You have to ask yourself how much is self-inflicted? How much is due to the market? Then you have to quickly bring your cash flow and your head count in line with the new reality. And then you ask, “What plays are we going to run through this transition, and how will these position us for the future?” If you were investing in too many things before, shrink back down where you are. Paint your North Star. Make one big bet."
- Beyond the hype - How to use chatGPT to create value - Strong post by Thomas Wieberneit on enterprise uses of ChatGPT - though ChatGPT is not the only generative AI tool enterprises could consider. My only beef? I too like the enterprise search use case, but I'm not convinced it's as "short-term" ready as Wieberneit says. ChatGPT is trained by periodically ingesting large data sets (its current data set is not current). Can it now incorporate short bursts of fresh enterprise data in near-real-time? This needs digging - but Wieberneit's use cases are interesting.
- ChatGPT for Industry Research: Not Ready for Prime Time - Another solid enterprise-related ChatGPT post, where Frank Scavo puts ChatGPT to the test, and contemplates how it can help in research - and where it falls short. I left a monster comment, calling out LLM (Large Language Model) cheerleaders for overstating what is possible with LLMs.
- 7 problems facing Bing, Bard, and the future of AI search - And so it begins - for background.
- Microsoft “lobotomized” AI-powered Bing Chat, and its fans aren’t happy - So, after Bing (and some gullible journalists) got worked up over Bing's trained-by-the-Internet-cesspool "Sydney" bot, some of its capabilities were curtailed (including long conversations). That led to yesterday's about face: Microsoft now says AI-powered Bing chatbot is 'not a replacement or substitute for the search engine.'
To which I say:
Microsoft now says AI-powered Bing chatbot is ‘not a replacement or substitute for the search engine’ https://t.co/5XZR4SYbV5
->a stark but hardly unsurprising about-face. Anyone who delves into LLMs can quickly figure out that consumer search is a terrible/not-viable use case. pic.twitter.com/ZBOwbOu2tR
— Jon Reed (@jonerp) February 20, 2023
Those who "win" with these new AI tools are going to need a very exacting standard of what works and what doesn't - so let's put down the AI pom poms and get to it, shall we?
Here's my security whiffs rule: if you get breached multiple times over years, I get to take a cheap shot:
GoDaddy says a multi-year breach hijacked customer websites and accounts https://t.co/HtZ1yVoxWE
-> maybe changing names to OwnedDaddy may be in order?
— Jon Reed (@jonerp) February 20, 2023
So... cheating on an AI ethics exam with AI - is this what pundits call the "circular economy"?
A Student Used ChatGPT to Cheat in an AI Ethics Class https://t.co/iCntWEWyl0
-> hey, the goal is to pass the class. actually being ethical was never the point....
— Jon Reed (@jonerp) February 18, 2023
Finally, I asked Brian Sommer if he tried out the "gift of backups" for Valentine's Day:
hey, it's not very romantic, but we all could use a backup. :) pic.twitter.com/520Wxa13yF
— Jon Reed (@jonerp) February 13, 2023
It looks like Brian wisely went in another direction:
the warmth, sincerity and heartfelt affection we all should have for our backup software IS the definition of Valentine's Day. Isn't it???
If I gave this to my wife, it would be rolling pin time!
— Brian Sommer (@BrianSSommer) February 13, 2023
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.