Enterprise hits and misses - Adobe and Figma fall out, the EU AI Act ...

Lead story - Adobe's acquisition of Figma falls through - what's the fallout?

As we veered into the alluring respite of the holiday season, the enterprise news cycle did its utmost to pull us back in. Not the least of which was Adobe/Figma. So, what happened? Stuart parsed the news in UK and EU kill Adobe and Figma's $20 billion takeover deal - what now?

Interestingly, Adobe explicitly blames non-US regulators for the collapse of the deal, arguing:

'There is no clear path to receive necessary regulatory approvals from the European Commission and the UK Competition and Markets Authority.'

But as Stuart notes, it wasn't smooth sailing in the US either:

Last week it was also revealed that Adobe is being investigated by the US Federal Trade Commission over its subscription cancellation policies, a probe that the firm admits may leave it exposed to “significant monetary costs or penalties and could have a material impact on our financial results and operations."

Exactly what knocked over the dominoes may never be known:

Given how long the approval process has rumbled on and given that final decisions were only a couple of months away, it's intriguing to wonder what has occurred to convince both firms that they were on a losing streak?

This would have been an interesting test of the current US regulatory environment, but that test will have to wait. Stuart speculates that someone will have to account for this mishap - will an executive or two be compelled to parachute out?

But the most interesting part is: where the two firms go from here. There was a reason for this commercial marriage. What does that say about the vulnerabilities of each party? On the Adobe side, where will capturing the imagination (and adoption) of the next generation of designers go from here?

The diginomica year in review extravaganza

Before the diginomica team fans out for eggnog and arguing with relatives about AI roasted chestnuts, we craft our opinionated reviews of 2023. You can see the full run down here. I'll pick a few for highlights here, but we cover everything from the year in marketing (Barb), to the year in Frictionless Enterprise (Phil), to the year in sustainablity (Madeline).

Derek covered off the year in enterprise use cases, Alex did the year in thought leadership from our partners. Meanwhile, I reviewed the year in AI use cases (including our first live gen AI project stories), and I co-authored the year in cloud ERP with Brian Sommer (one LinkedIn reader described our ERP review as "brutal"; I didn't find it that harsh but I'm taking it as a compliment).

A few more year in review picks:

HR tech - Brian boiled down five thousand HR tech briefings into 2023 - the year in HR Tech: "When it came to generative AI solutions, I saw lots of mimicry and incrementalism, but not a lot of oh-wow innovation." Retail - as the retail year surged to a holiday fever pitch, Stuart contrasted this year's winners and losers in 2023 - the year in Retail: "The COVID-driven mass movement to the internet faded as shoppers got back into the idea of physical retail. That led to retailers of all shapes and sizes having to rethink their efforts to strike an omni-channel balance that works for them and for their customers. Meanwhile generative AI is going to solve all our problems...apparently. " AI policy and robotics - Chris had the unenviable challenge of capturing 2023's AI policy skirmishes in just ten articles (2023 - the year in AI Policy): "Less than half (48%) of organizations have put in place policies for responsible AI adoption – down from 52% in March." Also see: 2023 - the year in Robotics.

Jon's grab bag - it was an AI-overdose kind of year, so it figures that the final pieces of the year would be AI-related. Katy assessed another AI-odd-couples deal in Meta and IBM hold hands to launch the AI Alliance. Martin got my dander up a bit with AI without training? It can be done, argues Appian CEO Matt Calkins - but it seems we don't disagree that much in terms of the central role of quality data in all of this. Oh, and that the enterprise has a chance to get a better AI result than we've seen with the might-as-well-train-it-on-Reddit-too  LLMs on the consumer tech side.

Neil made a persuasive case for open source AI (and why regulations could hurt open source) in The EU AI Act - dissecting the detail. Finally, George wrapped up a year of strong AI analysis as well with New practices needed for trustworthy AI - some expert commentary. Given how fast AI models arrive/evolve, I find the possibility of "embedding relevant standards and practices directly into AI processes in an actionable way" an intriguing option. Stay tuned...

Best of the enterprise web

My top eight

What is the 2023 economic outlook? Consumers and CXOs give clues

US retail sales rose 3.1% this season -- down from 7.6% last year - As of press time, I wasn't able to find brilliant retail analysis from this holiday season. But I did find this useful summary from The New York Post:

Arun Sundaram, an analyst at CRFA Research, said many shoppers waited for Black Friday and Cyber Monday to make holiday purchases and finished the final sprint during Super Saturday — the last shopping day before Christmas. 'Consumers are still spending, but they’re still price conscious and want to stretch their budgets,' Sundaram said... Online shopping accounted for a large chunk of this year’s holiday spending. According to the Mastercard report, online retail sales jumped by 6.3% year-over-year, while in-store sales rose just 2.2%.

What does this mean for the economic outlook in 2023? This article implies that things may be looking up, but just a tad:

Americans have been saddled with soaring prices in recent years though there have been signs inflation is beginning to cool. The Fed’s preferred measure of inflation — the Personal Consumption Expenditures price index (PCE) — rose less than expected in November.

Could this subsequently (finally) lead to lowered interest rates again? Time will tell. Constellation's Ray Wang noted a similarly optimistic dynamic from enterprise CXOs:

To conclude the 'CxO Business Confidence Survey,' executives were asked whether their enterpraise was experiencing a better business climate in 2023 than in 2022. The question elicited an optimistic response of 'yes' from 51% of the CxO respondents. This has shifted from Constellation’s 2022 Q4 CxO survey, where 67% predicted a worse business climate for 2023 overall (see Figure 8 - Survey: 2023 H2 CxO Business Confidence Survey).

Will Open source win over proprietary LLM for enterprises? Vijay Vijayasankar tees up one of the most important issues of enterprise AI in 2024: "The enterprise world is one full of narrow use cases. What’s important there are things like accuracy, compliance , security and so on. While enterprises will absolutely benefit from multi modal AI – there is plenty of value that can be extracted with just text based models. Given that case – I would think that an open source model that is fine tuned (and most probably with RAG ) is the more viable approach for most companies. Also given GDPR like use cases – they will need work arounds on how to implement things like 'right to forget', 'data localization' etc." I left a lengthy comment, summarized here:

Those pros/cons included flaws in open source training models, IP and safety risks in those models, and the difficulty of creating guardrails (in theory, open source can be superior in those aspects as well, especially if the model training is transparent). Which leads us to this:

AI image training dataset found to include child sexual abuse imagery - The Verge reports on an important research study from Stanford, with big implications for open source AI: "The researchers acknowledged it would be difficult to fully remove the problematic content, especially from the AI models trained on it. They recommended that models trained on LAION-5B, such as Stable Diffusion 1.5, 'should be deprecated and distribution ceased where feasible.' Google released a new version of Imagen but has not made public which dataset it trained on aside from not using LAION." More on this story: Exposing the Rotten Reality of AI Training Data (podcast and transcript) Social Security numbers of some Xfinity customers vulnerable in latest data breach: What to know - aside from AI training date controversies and retail numbers, the other big story of the holiday season was damaging cyberattacks, based on clever new strategies: "And these hackers are really pernicious. They didn’t hack into Xfinity per se, but they implanted their malware into [Citrix] software that Xfinity bought. Weisman says these are called “supply chain” hacks and they are a growing problem." Sidenote: the Citrix vulnerability story isn't over yet either. Also see: Louis Columbus on Why attackers love to target misconfigured clouds and phones. Tesla Updating Autopilot In Nearly Every Car In U.S. After Investigation Into Crashes - challenges/obstacles/breakthroughs in autonomous vehicles will be another story to watch in 2023. I suspect we won't see big tech breakthroughs in this area in 2023, but we may see progress in finding more controlled environments where "autopilot" can be successful (e.g. parking garages, shop floors). Things are about to get a lot worse for Generative AI - Gary Marcus previews research that lines up well with the current New York Times IP lawsuit against OpenAI and Microsoft: "OpenAI’s image software (which we accessed through Bing) is perfectly capable of verbatim and near-verbatim repetition of sources as well. Dall-E already has one minor safeguard in place – proper names (and hence deliberate infringement attempts) reportedly sometimes get blocked – but those safeguards aren’t fully reliable." Three Decades of Software Vendor Selection - Frank Scavo continues his fruitful series of project reflections, including customization debates that never seem to get old.

Oh, and this uncensored wrap show was fun:

Whiffs

It seems like I've covered a number of whiffy situations already, but wait - there's more. Via Brian Sommer: A Chevrolet dealer offered an AI chatbot on its website. It told customers to buy a Ford: "It's not every day that a Chevy AI program heaps praise on an archrival, the Ford F-150."

Speaking of Scavo, he had a fun holiday run-in with a very persistent PR person:

That inspired me to update a 2023 piece on classic PR pitfalls - and how to avoid them.

Oh - and thanks for your loyal readership year in, year out. Happy new year and 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.