Tuesday, 25 June, 2024 / Edition 13

Y’all, I finally caved and put some ice in my instant coffee for instant iced coffee. Iced instant coffee? Anyway…between the “super concentrated instant coffee” step and the “add ice” step, I also added my chocolate protein shake. So, it’s a bit of an instant iced mocha (I’m committing to that word order). Because my expectations were low to begin with, I have to admit I’m not disappointed.

I needed some good go-juice to keep the productivity and buzz going after a fantastic trip to Houston last week for URTeC 2024. I appreciate everyone who stopped by the booth to say hi! Predictably, AI/ML and automation frequently dominated the technical and innovation sessions. I’ll dive into those in the coming weeks, but as a warmup this week, I’ll review how AI is not so gently making its way into our everyday lives, as well as how fault network complexity can impact its method of transmitting stress. Let’s do it.

Sarah Compton

Editor, Enspired

AI is Probably in Your Pocket

Patiwat Wichayapakon/Shutterstock.com

AI is coming for…I mean to…your phones.

Catch up fast: Apple announced its Apple Intelligence about a week ago, but it’s a bit late to the party (not hard to do in the AI/ML world! You sneeze and you’re late), as several other phones like the Samsung Galaxy S24 Ultra and Google Pixel 8 Pro have AI baked in already.

Why it matters: Individual employees increasingly find themselves to be the target of bad actors hoping to breach their employer’s cybersecurity. It’s imperative we understand the risks and rewards accompanying the new technology we’re all about to carry in our pockets.

New features: I am alternating between, “Take my money,” and “Hard pass,” with some of these capabilities, which include:

• Custom emojis! I am more excited about this than I should be. Apple Intelligence gave an example of creating a dinosaur on a surfboard wearing a tutu, and I have several instances in which I could use just that emoji.

• Meeting notes and e-mail summaries. There are apps for this, but now it’ll be a feature that comes with the phone automatically. One less thing to download and click on? I’m in.

• Coordinating your schedules. Just as a random and unrelated example, let’s say you have a consultancy, a full-time job, and a family with kids and a husband (again…totally random and definitely not about me! 😉). That’s a lot of schedules to keep track of, and sometimes coordinated calendars just don’t cut it. AI has the ability to read your e-mails and maybe your texts to then coordinate your calendars in case you forget to track activities. You got an evite for a kids’ party next weekend? Warning! You also said you’d be interested in camping. No more double-booking!

Potentially problematic: The ability to summarize key points and coordinate schedules makes my soul happy, but “read your e-mails and maybe texts to extract information” makes my skin crawl and sets off alarm bells in my head—especially considering the highly sensitive and proprietary information exchanged at times.

The bottom line: If AI is isolated on the phone and won’t send information into the ether, then maybe it’s OK, but the potential for a breach here seems huge.

A message from TGS

This whitepaper features a case study that demonstrates the use of an automated stratigraphic correlation workflow, specifically the ChronoLog pipeline, to develop a geologic model of the Midland Basin.

Key points include:

ChronoLog Pipeline Usage: Employed to create a geologic model of the Midland Basin’s subsurface geology.

Enhanced Accuracy and Efficiency: Improved the construction of 3D stratigraphic and property models.
Widespread Application: The resulting interpretation is being evaluated and used in various workflows.
3D Geologic Volumes: These volumes are utilized for well planning, geosteering, detailed reservoir studies, and simulations.

Fault Network Complexity Affects Slip Rates

Oliverdelahaye/Shutterstock.com

A recently published Nature paper explains the relationship between fault complexity and the rate at which slip happens.

Breaking it down:

• Fault network complexity is considered higher when faults intersect each other at high angles.

• Networks with lower complexity accommodate stress through creep, while higher complexity networks are more likely to stick, build up stress, then release it in bigger seismic events.

• The authors say findings encourage building more complexity into computer models to simulate the process more realistically.

The goal: This study worked to see how natural fault networks respond to stress exerted over long periods of time. Typically, the industry’s goal is to understand the fracture networks created by nearly instantaneous large stress perturbations.

Why it matters: Fracture models in the industry often over-simplify in favor of computation time and data requirements for initial conditions, among other factors.

• Some models work to build more complexity in the natural system through discrete fracture networks, but DFNs require inputs (i.e. measurements) for their generation. The industry is often hesitant to spend the money needed to gather these data, because the payoff isn’t always clear.

• Pseudo 3D and even blatantly 2D models have been producing results for more than a decade, and sometimes those results are highly useful.

Key takeaways: Trade-offs between improved accuracy, logistics of the model build, and economic usefulness must be made. The key here is to know the question you’re looking to answer, understand the full implications of model assumptions, and interpret results within that context.

Go deeper: The full paper can be found here, and a good summary is here.

👍 If you enjoyed this edition of Enspired, consider supporting AAPG's brand of newsletters by forwarding to a friend or colleague and signing up for our other newsletters here.

➡️ Was this email forwarded to you? Sign up for Enspired here.

✉️ To get in touch with Sarah, send an email to editorial@aapg.org.