AI vs. IQ, who will win?

Artificial Intelligence (AI) is one of the most exciting technological advancements in recent years. I not only use it for my work, but I have also integrated AI into my personal life to the point where my wife rolls her eyes at me. AI is indeed a significant development, but if no one has told you, do not let it discourage you from thinking for yourself.

Make no mistake, I am not an AI denier. AI can be used to accomplish tasks in seconds that would take me hours, days, weeks, or more. For example, take supply chain risk modeling and risk mitigation. I, or some nerd like me, can go through a series of specific processes to identify risks, assess risks, mitigate risks, and plan for a program to continue the journey of bulletproofing a supply chain from risks and achieving supply chain resiliency and sustainability. But as those of us in the Southern United States say, ‘it may take a minute’. With AI, once all the current state data, direction, parameters, and defined objectives, to name a few, are in place, in a matter of seconds, AI can churn out the structure for a risk mitigation plan.

Here is my ‘yes, but’… Yes, but don’t in any way, shape, or form consider the AI-born plan a plan. Consider the AI result as a starting point to start using your own brain, not a tool to stop using your brain.

As a test, I asked the AI to route full container loads of palletized pickles in pickling juice, packaged in glass jars, from Chicago, IL, to Oslo, Norway, identifying the most effective modes of transport and equipment to complete the task. Below is the result, and AI did a darn good job at producing an initial plan for me to ship my pickles.

AI-Generated Shipping Plan

Best Mode of Transport

For full container loads (FCL) of palletized glass jars:

  • Ocean Freight is the most cost-effective and practical mode. Route: Chicago → East Coast port (e.g., New York or Norfolk) → Transatlantic shipping to Oslo Transit Time: ~20–30 days depending on port schedules and customs clearance
  • Rail + Trucking: Used domestically to move the container from Chicago to the port of departure

Air freight is generally not recommended for this commodity due to:

  • High cost
  • Weight and fragility of glass jars
  • Lower urgency for shelf-stable goods like pickles

📦 Equipment & Packaging

To ensure safe and efficient transport:

  • Standard 20-ft or 40-ft Dry Container 20-ft: ~10–11 Euro pallets 40-ft: ~21–25 Euro pallets
  • Palletized Loading is preferred over floor-loading for fragile goods like glass jars Easier handling with forklifts Reduces breakage risk Allows for shrink-wrapping and corner protection
  • Use Euro Pallets or Standard Pallets Ensure jars are tightly packed and cushioned Avoid overhang to prevent damage during loading/unloading
  • Void Fillers & Dunnage Use honeycomb paper, foam, or airbags to prevent shifting
  • Labeling & Documentation
  • Mark as “Fragile” and “This Side Up”
  • Include customs paperwork and food safety certifications

📅 Best Time to Ship

  • Avoid peak seasons: Late Q3 and Q4 (holiday rush) can drive up freight costs and cause delays
  • Ideal shipping windows:
  • Spring (March–May): Lower congestion
  •  
  • Early Fall (September): Before holiday season spikes

End of AI Transport Plan

After receiving the AI results for my pickle shipments, I examined the end product and played a little game of “I Spy.” Here is what I spied.

AI did a good job picking modes of transport. Truck-Rail-Ocean fits the need for practical and economic purposes.

AI also noted that the pickles are in glass jars, so packaging needed to be stout with plenty of void fillers and dunnage used. Well done so far.

I was impressed that AI considers holding off on shipping until off-peak season to save a shipping buck, but for many business models, that isn’t very practical. Most pickle fans expect to be able to eat pickles year-round, and sellers most often look for ways to reduce warehousing costs, not increase them.

Okay, let us take this transport plan to our pickle customer, ‘the pickle boss’, and get some pickles booked for transport, right?  Not yet, we have to make one adjustment to the AI shipping plan.

AI selected 20’ and 40’ standard containers as the best equipment for shipping our pickles.  There is one factor AI did not consider when choosing regular dry containers. In the winter months, it gets remarkably cold in Chicago, IL, as well as the path to Newark/Elizabeth, which is the likely port of exit. Not to mention Norway, along with all of its neighbors, is a notoriously cold country.  Depending on the time of year, the question would not be whether the jars of pickles would freeze, but rather at what point in transit the commodity would freeze. Once the pickles freeze, the product would be ‘mushy’, and no one likes a mushy pickle, so the commodity would not be salable.

Moreover, since most pickles are packed in liquid and glass jars don’t expand, the good people of Oslo, Norway, would have a huge number of cracked pickle jars, as well as pickle juice, on their hands (pun intended).  It is clear AI should have mentioned using refrigerated containers for at least half the year to maintain the interior temperatures of the container above freezing.

For the record, chances are normal marine cargo insurance would not pay a claim for the shipments, so the shipper would be in a pickle (Yes, I know, I couldn’t help myself). Since cold ambient temperatures are a regular occurrence in shipping, the insurer, as a condition of insurance, would require the commodity to be packaged and shipped to withstand the rigors of normal transportation. In short, no reefer containers, no payable claim.

Do frozen, leaking, or cracked jars of pickles make AI not smart to use? Of course not, AI is too productive a tool not to use. The lesson of the pickle is not to let AI overrule your brain. Your risk management team should utilize AI for your supply chain risk mitigation efforts; however, your supply chain risk management efforts should not be AI. Even though AI gives us access to virtually every bit of information on the internet, the application must be correct. AI results are not always relevant to our nuanced needs, and the ‘ask’ must be incredibly specific, methodical, and inclusive for AI to work as designed. Second, as humans, if we stop doing things because AI can do it for us, before too long, we won’t be able to do things.

I am not speaking of ‘things’ as the same as our younger generation not knowing how to dial a rotary phone, operate an IBM Selectric typewriter (ask your parents), or read a paper road map; those things are obsolete, and young people have no need to know them. This is about continuing to use our thinking and creativity skills to their fullest potential, enabling us to keep them and not delegating the job out.  Use AI to complement and help execute your thinking, not replace it.