You guys. I am so sorry. I know I write about this stuff all the time. I JUST CAN’T GET ENOUGH. NLP. GPT-3. DALL-E 2. GPT-3 again. DeepMind. Nobody is ready for how deeply AI tools are going to penetrate into artists’ creative processes.
Check out this showcase of results from a tool (available for public use) released by Midjourney, a research lab focused on “expanding the imaginative powers of the human species”.
This image was created from nothing but the text “bronze statue of henry cavill, photography, beautiful, HDR, 8K, 80mm lens”. LOOK AT THAT. Is that the same thing as a human creating a wholly original work of art by hand? No. But often I don’t want to make the picture – I want to use the picture to illustrate something. And in that case, simple text prompts that generate amazing images that match the concepts in the text are absolutelygood enough for me. PM
No, this is not about a bot that prevents the monthly “extract Excel from ABAP” blog posts and redirects to ABAP2XLSX instead, sorry!
We’ve written about the SAP Community calls before (e.g. Event Mesh and SAP CAP). These are usually covering cutting-edgy subjects with less marketing speak and are very helpful. But lately I’ve not heard of any calls and was wondering what’s up.
Turns out that instead of visiting the SAP Community page, now we need to sign up for the upcoming calls on YouTube channel. Community calls are listed as ‘Upcoming live streams’ (look for the pink-orange background). The upcoming call on August 31st will be about AI-Powered Personalized Recommendation Service, which sounds pretty cool.
Continuing AI theme, there are also 2-minute videos about different SAP AI offerings. I find the Document Information Extraction service particularly interesting and recommend exploring it at your leisure. JP
Enterprise software development is not just about algorithms and programming languages. It’s also about business processes. And for every enterprise, this means accounting.
It’s always weird hearing the accountants talk. “Clearly, this chargeback goes to a revenue recognition account. – Indeed. But we need to offset tax liability for asset depreciation. – Quite!” I imagine that’s also how accountants feel when developers are discussing OData filters.
One does not need to go back to college though to learn just enough about accounting to understand that the dialog above is pure nonsense. Recent article Accounting for Developers explains accounting basics very clearly and is reasonably short for the volume of information it covers. I highly recommend it to all the SAP developers in particular. (As we know, all roads in SAP lead to FI.)
I feel like every other issue I’m throwing something out from Bob Evans. Apologies, Bob – maybe if you didn’t write such interesting stuff I wouldn’t regurgitate and link to it so often. I’ve written about industry cloud before, but this time Evans takes us to it through the lens of some of the older enterprise players: Oracle, SAP, and IBM.
With Oracle, Evans calls out a recent huge acquisition, industry-specific apps for more than a decade, and positioning industry as central to Oracle’s business. I’ll add that Oracle’s impressive catalog is what sticks out in my head as its strength. SAP gets kudos for half a century of industry insights and RISE being a success (I wish I knew more details on that). From my view, SAP should by all rights have the largest store of industry brainpower. IBM gets praise for unifying its consulting with its product, and I just gotta agree with that. Is there any other player that can match firepower with boots on the ground at that scale?
I don’t even know if it’s possible to be too optimistic about cloud. From infrastructure to platform and software, it’s just gonna keep steamrolling through and changing the equation on everything. Until the next thing. PM
In a recent LinkedIn post, SAP integration expert Michal Krawczyk inquired what do others think about an EDI myth “APIs are better suited than EDI for small to medium enterprises”. As an IDoc book author, of course I have an opinion to share.
This is not really about the enterprise size. “EDI vs API” is not even apples vs oranges, it’s more like HTTP vs bananas. While both apples and oranges are fruit, EDI and API are different species.
EDI is a standard format for business information exchange. It signifies agreement between everyone involved that specific data would have specific format, place, and meaning. One might say that API kind of does the same thing (it tells you what data in what format it needs). But API is not a standard per se, while EDI has specific global standards. Also, API is more about the means of information exchange, it’s about functionality itself.
You can even use APIs, if you wish, to implement an interface based on an EDI standard (heck, you can even blend IDocs and APIs). So “EDI vs API” is just not the thing for any enterprise size. Myth busted! JP
“SAP” and “future” are concepts that rarely connect in my head. That’s not meant to be a hater comment – just that the sci-fi fanboy part of my brain thinks of the future as the place where we travel to the stars, not where we match invoices 14% better. SAP does actually employ futurists, though. This blog caught my eye and got me thinking. It’s a profile of Upen Barve, who along with Martin Wezowski is co-founder of The Future Hub at SAP.
“We spearhead & shape SAP’s long-term innovation vision, produce and supervise future concepts, create Macro POV’s and build concept cases and concept-prototypes to make the vision tangible. We inspire & strategize with our ecosystem to investigate possible futures, and validate with tangible PoCs.”
Upen Barve on Future Hub
It takes a bit for the interview to get going. Lots of high-sounding phrases that – while I’m sure have lots of meaning to the futurist crew – don’t take shape for me as a reader until he goes into detail on a few AI outputs and predictions:
Barve’s four pillars of human ingenuity and machine intelligence coming together: individual augmentation, enterprise decision augmentation, autonomous operations, autonomous networks
Points to a forecast saying between 2042 and 2045 that machines and humans will be at equal capability. Calls it “inevitable”.
Will AI be able to learn emotions? He says gut feelings will continue to distinguish humans from machines.
I suggest going out to Barve’s LinkedIn profile and looking at some of the videos to get further details. While the future is necessarily murky, futurists and fusionists alike should work to be clear in their communication. PM
Over at diginomica, Mark Chillingworth describes a few CIOs out there in the world doing the best they can in the tech talent crunch. They’re making headway by transforming recruiting from an externally-sourced job to an arm of IT itself. A key finding:
talent acquisition is, like technology, not just a bolt-on, but a strategic approach and a good investment.
The folks Chillingworth talked with saw benefit from in-sourcing recruitment. Vital to success appears to be giving recruiters incentives to go the extra mile, and relationships that can only be built with colleagues. When tech talent ups its game, the business notices.
I often work with IT teams. It is no lie to say they are stressed, short-staffed, and in need of skill-sharpening time. Getting good tech talent means investing in lots of things, and recruiting is a key factor. The best people are never on the market, so you need something more than LinkedIn spammers to find them. I have worked with recruiters in the past and I can tell you: if the recruiter is invested in the business they’re hiring for, it makes a huge difference. Recruiting follows the same adage as other talent: you get what you pay for. PM
I don’t know how I got there, but I landed on the SAP Open Connectors help portal. Open Connectors helps “unify the developer experience across all kinds of applications and services”. Check out how many connectors are available for integration use.
Every platform I touch has a similar concept – there’s a need for simplifying connections between apps, clouds, databases, and everything in between. If you can draw a picture of what you’re trying to do, you can probably find a pre-built connector somewhere to smooth the path.
The ubiquity of connectors shines a light on that eternal question: build or buy? If I can make a Visio drawing and just plug in little pieces with API keys and credentials, doesn’t that lead to agility? If I can hand-write my own custom application in any language I choose, doesn’t that mean that I can truly make anything I desire?
If I need to build something in order to communicate that thing’s value, I will glue connectors together all day long. If I need to make something that can stand the test of time, I’ll happily open VSCode. But the world is changing, and I don’t think that distinction will remain as clear as it seems at the moment. PM
A corporate BS buzzword focused on by companies who don’t want to lose their Millennial/Gen Z workforce, mumbling something about making safe communities and locally resilient supply chains.
An environmentalist term meaning this object or activity can persist in such a way as not to immediately destroy our planet/climate.
A couple of items in the last week made me think about these two definitions. First, Julia White of SAP gave a quick talk at Cloud Wars Expo which included some mentions of sustainability. I am cynically inclined to interpret most corporate folks as acting in the first parsing above – but Julia and her interviewer’s chat actually left me feeling more charitable. Give it a listen.
With regard to the second, I saw news that Google and Oracle suffered data center outages in a recent London heat wave. The cloud computing market sure ain’t gonna get smaller, so cloud providers have to both plan for heat disasters and – more importantly – be good citizens of power usage. A study in Science found that data centers use around 1% of all worldwide electrical power, and a great deal of that power goes into cooling. Climate change should be top of mind for an industry that relies on cooler temperatures and massive amounts of power.
Time will tell whether corporations are truly serious about buzzword-chasing. I hope they are. PM
In the beginning, there was the big, bad monolith that did not want to change. Then along came Netflix and said let’s break the monolith apart into the small pieces. These pieces could then be put into the containers and shipped off to the Cloud to be easily scaled up or down, depending on the demand.
These Lego brick-like pieces called “microservices” and they are supposed to make it as easy to scale SAP systems as to open more cash registers at a supermarket. Except when not.
Turns out monolith is not necessarily bad and microservices are not always great. In this interview, the author of Monolith To Microservices book Sam Newman talks not just about how to break up a monolith but also when you probably shouldn’t. This quote could not have been truer in the SAP world that is sometimes too tech-obsessed after being tech-starved:
We focus on the tech tool, not the thing that the tech tool lets you do.
When considering microservices for your next project, think what they would “let you do” and if they’re the right solution in the first place. To avoid, you know, a megadisaster. JP