This article is approximately 398 words long. Click here if you would like to request a complimentary copy.
This article is approximately 195 words long. Click here if you would like to request a complimentary copy.
A book that gave me a lot of pleasure some time ago is M.J. Moroney’s Facts from Figures. I seem to have lost my copy and it is not full-text on Google but one term that has stayed with me is ‘Snibbo’. Snibbo fist appeared in the early 20th Century Daily Express cartoon ‘Beachcomber’ and referred to a mythical product that among its many other virtues, ‘killed greenfly and stopped cliff erosion’. Moroney repurposed Snibbo to introduce statistics as a tool for verifying improbable claims from the marketing department. The idea that folks pumped out dubious claims for their products and services has stuck with me. I am constantly looking for Snibbo. And, I’m afraid, finding it everywhere. So I’ll be seeking out the Snibbo this recap of our output from the last year.
In our first edition we reported on a US National Academy of Science analysis of the ‘Digital Twin’. Seeking Snibbo sometimes requires reading between the lines. Not in this case. NAP reveals the DT to be essentially Snibbo viz. ‘the publicity around digital twins and digital twin solutions currently outweighs the evidence base of success’. And again ‘it is challenging to separate what is true from what is merely aspirational, due to a lack of agreement across domains and sectors as well as misinformation’. In the same issue we fired our first shot across OSDU’s bows asking ‘where are the presentations? Where are the standards?’ Like other ‘standards’ initiatives, The Open Group is in a bind when it comes to making its material freely available. If it’s free why should anyone pay to join? On the other hand, if it’s ‘open’ … err… This paradox is a long term issue with oil and gas joint ventures which often hide behind the idea that there is intellectual property to protect. There is another explanation, that the presentations are not up to much and that publishing them would expose the Snibbo.
An attempt to de-Snibbo the data discourse came from the UK/Norwegian Society for Professional Data Managers which we have reported on since its inception. It was a good formula, knowledgeable people, timing that fitted in with a daily work schedule. What could go wrong? Unfortunately SPDM was wound up late this year citing a decline in membership. Perhaps not enough Snibbo for ‘sustainability’?
We like the Dutch USPI-NL standards body for its openness. Its presentations are online and visible to all. It also has an broad purview spanning oil and gas, chemicals and nuclear. USPI-NL, having been deprived of its CFIHOS flagship a couple of years back, has retaliated with FL3DMS, the Facility Lifecycle 3D Model Standard a.k.a. a ‘foundation for the digital twin’. More Snibbo?
Our second edition of 2024 featured our first report from the EAGE Digital conference where we led with BP’s presentation on AI in seismic interpretation. Frequently the problem with oil company presentations is that there is not enough Snibbo. Or rather that oils prefer to emphasize their in-house expertise as opposed to what is provided by the software supplier. Such omissions are actually encouraged by the conference organizers wishing to avoid ‘commercialization’. I usually try to mitigate this by asking ‘what software was used’. Back in the day this was like pulling teeth. Folks are more open these days fortunately.
Our report on the TietoEvry hack led us down a rabbit hole to the Snibbo-free Firewall Times that reported on multiple breaches of various cloud services. The website has not been updated for a while possibly showing that Snibbo-free business is hard (we know!) My editorial on shorting NVIDIA proved rather poorly timed. Since it went online, NVIDIA’s share price has risen 50%. Having said that, my Snibbo-oriented view of the world leads me to believe that NVIDIA and the whole AI bonanza will end in tears, eventually. Palantir, another stock market darling and Snibbo major, burst onto the upstream scene as OEDA came out of left field as a competitor to OSDU.
N° 3 saw some negative Snibbo as a group of process control vendors were heard ‘trash talking PI’. Our words not theirs of course. But there is an argument to be made that when an industry staple gets bought by a conglomerate it loses some of its pizzaz. Our peek into quantum computing suggested that its early Snibbo/promise has yet to be fulfilled unless quantum annealing, quantum communications or Quantum of Solace are key to your business. More reporting from EAGE Digital investigated two ‘disruptors’ – AI and OSDU. Both of which have indeed disrupted the daily work patterns of the major oils. Step back all you geo/engineering domain specialists and make way for the Pythonistas and crafters of retrieval augmented queries. Having said that, our EAGE Digital report in issue 4 is definitely worth a read or re-read if you did not see it before. En passant we noted a paradox from the AI/LLM brigade that on the one hand, it will revolutionize work but on the other hand, ‘AI will never replace someone’. The EAGE Digital offered our first in-depth(ish) examination of OSDU in the wild with rather underwhelming results. Shell observed that OSDU’s ‘Frankenstein’ code base ‘needs attention!’ For Halliburton, ‘companies have ‘totally underestimated the cost (of OSDU)’. Our report from Norway’s Force conference on risk management demonstrated that wildly optimistic forecasts of production and reserves is widespread. We were delighted to find a polite way of referring to such in-house Snibbo with the wonderfully euphemistic ‘strategic misrepresentation’. We were also interested to learn that AI is being driven from the top. Chevron’s chairman for instance, thinks it will ‘transform every part of the value chain’. TotalEnergies has set up an AI factory with hundreds of developers. Some are training the ‘citizen developers’ although not all agree, ‘not every use case merits an app. The number of people in BP who need to do pressure transient analysis can be counted on the fingers of one hand’.
N° 5 saw our bombshell report on ‘turmoil’ in the OSDU Forum and the realization that ‘OSDU currently maps ‘every data type in the world’ but comes with ‘no functionality’. Actually this was anticipated in our report on OSDU’s ‘Zero end users’ back in 2023. I enjoyed writing my editorial on ‘OSDU - why no data egress?’ and I understand that quite a few readers did too. Also the discovery of the DARPA Heimler Catechism was a gem. If you ever want to demolish a project, just run it through Heimler! Having said that I’m not sure if the army takes its own advice, witness the Army’s $435 hammer unveiled by the 1986 Packard Commission. Our review of Blockchain in oil & gas is also worth a re-read. Blockchain has surfed on a sea of Snibbo which has hidden much of its logical inconsistencies. No it can’t guarantee the validity of anything outside of its chain of numerical tokens, not even with an ‘Oracle’. 2024 saw the re-emergence of interest in semantics with a new ‘IDO’ replacing ISO 16926. The 20 year old RDF technology is now being associated with up-to-the-minute LLMs. Brazilian researchers from UFRGS have demonstrated an ‘ontology-driven digital twin’. Snibbo on top of Snibbo?
In this issue, N° 6, we give CFIHOS another shot, our best yet. We were pleased to note that CFIHOS is inviting submissions of ‘conformant’ software. While we would have preferred ‘compliant’ (a no-no for most upstream orgs by the way), the distinction is somewhat irrelevant as at closing time there were exactly zero ‘submissions’. We were also intrigued by the ‘frenemy’ relationship between CFIHOS and DEXPI. Rather like that between CFIHOS and the CII AWP offering.
You may be thinking that Snibbo is just the equivalent of marketing hype and you would be right. But a quick look at Google N-Gram finds precedence for ‘Snibbo’ (taking off in 1960) over ‘hype’ (1980). You might argue that I am seeing Snibbo in everything. Perhaps I should be kinder and less critical of folks who are invested in doing real stuff. I would like to be, but to judge from LinkedIn posts there are many, many more people involved in puffery, hype and … Snibbo, than there are folks trying to be objective. The hype has also infected the world of academic publication where Snibbo-like claims are re-hashed as facts. We are trying to fight back and raise the profile of Oil IT Journal by entering the research world with DOIs for our more informative pieces. If you are a publishing researcher and see something interesting in Oil IT Journal please do reference our DOI. It’s a bit like a LinkedIn ‘like’ but more ‘serious’!
Have a great 2025 y’all.
This article is approximately 1102 words long. Click here if you would like to request a complimentary copy.
This article is approximately 1998 words long. Click here if you would like to request a complimentary copy.
This article is approximately 219 words long. Click here if you would like to request a complimentary copy.
This article is approximately 1108 words long. Click here if you would like to request a complimentary copy.
This article is approximately 256 words long. Click here if you would like to request a complimentary copy.
This article is approximately 368 words long. Click here if you would like to request a complimentary copy.
This article is approximately 481 words long. Click here if you would like to request a complimentary copy.
This article is approximately 248 words long. Click here if you would like to request a complimentary copy.
This article is approximately 1185 words long. Click here if you would like to request a complimentary copy.
This article is approximately 259 words long. Click here if you would like to request a complimentary copy.
This article is approximately 225 words long. Click here if you would like to request a complimentary copy.
This article is approximately 881 words long. Click here if you would like to request a complimentary copy.
This article is approximately 1095 words long. Click here if you would like to request a complimentary copy.
This article is approximately 659 words long. Click here if you would like to request a complimentary copy.