David here. In the next 5 min, you’ll get:

💡What’s New: Mixed signals in daily markets with bright long-term prospects.

🤔 Opinion: The time is now to define your automation strategy.

🛠️ Tools & Data: New opensource ways to start generating more value.

Want to feature your service or product in DRIFFT? Apply here.

💡What’s new

After some feedback from the audience, I decided to shorten my weekly posts. From now on, I’ll provide just the most important topics from my week.

Economics. How fast will the electric energy transition be? 5-10-50 years? I am seeing conflicting signals from the resource markets and economists. On the one hand, many important battery metals futures indexes are sharply down from 2022. On the other, most projections reported in media show long-term supply shortages with strong consumer demand. FRED data series for US mining are all experiencing negative long-term trends or neutral growth (e.g., new business starts, employment, production costs). Combined with negatively trending global economies, I am in a wait-and-see pattern. I rather enjoyed a recent tweet by Señor Musk:

Bonus: here’s a list of “critical element” stocks to load into your favorite tracking application.

Tech. In May this year a symposium on GeoAI took place in which leaders from ESRI, NASA, AWS, and others met to discuss the future of geospatial AI. Key takeaways: (1) The earth-observation industry has troves of data waiting to be analyzed; (2) These organizations need opensource developers because the challenge is too large for even the biggest institutions; and (3) The industry must transition from simple surveillance to providing applications that affect a change. It’s coming.

In other news, there are many back-office tasks which are being automated using LLMs like GPT. The most prominent seems to be LangChain (or Llama Hub). These software kits can build AI agents, which are structured tasks handled by LLMs. Agents make it possible to automate many back-office functions, especially marketing and communications (which tend to be time and attention-intensive). This capability will have implications for venture funding. How so? Most new startups are expected to stay smaller for longer while they scale product (not headcount). Initial checks from VCs can therefore be smaller. In more capital intensive industries, like resources, it could mean lower barrier to entry. And with more competition can come more supply to market and lower prices.

Mining. One of the more expensive costs in any resource operation is power. It’s interesting to consider the profit impact that micro nuclear reactors–or any energy innovation–could have on mining operations, especially for remote areas that depend on costly diesel. Resource-rich western US states are exploring solutions that could turn uneconomic areas into major profit centers.

🤔 Opinion

We’re entering a golden age of hyper automation and personal freedom. Now is the time to define and begin implementing your (organization’s) strategy. And you should have one. Automation lets you free your time so you can focus on the important tasks of up-skilling and exploration. These will make you more productive in the digital economy.

I mentioned the LangChain framework above. In my opinion, task agents built around these frameworks have a bright future. Even if these AI agents don’t always get things right, it doesn’t matter. How often do all employees get things right? If your AI is effective at a single task 90% of the time and the human is 92%, is that really a big deal? Probably not for most simple tasks.

Decision systems are getting better. There will soon be fewer experts doing more management of AI-generated outputs. Organizational models, of course, will need to take this into account and adapt. It’s already happening in medicine and marketing.

Noah built before the flood. Be ready.

🛠️ Tools and Data

Share DRIFFT

Know someone looking to level-up their AI and opensource game in mining? Just copy and paste this link: https://drifft.beehiiv.com 

Thanks for reading! Want me to look into a particular topic? Email your suggestions and and I will dig.

David