Logistics is in BIGGER Trouble.

There’s been a truck driver shortage for almost two decades. I remember writing on the estimated shortage of 240K drivers back in 2013.

Moreover, with so many drivers being immigrants or cross-border drivers from Mexico, and the immigration crackdown in the US, it’s only become much worse, as chronicled yet again in the latest #HFSResearch piece.

However, I don’t think their answer of autonomous fleets in the answer. The tech isn’t there yet (as even Tesla can’t deliver fully reliable and safe autonomous vehicles yet, and it’s been working on them the longest in North America), half the states don’t even support testing of such vehicles yet, and, as always with new tech, we’re one bad accident away (as a result of rushed trials) from a major backlash that will stall progress for a decade.

I think it’s time we look back and take lessons from history (which I know most of my American colleagues have forgotten, or you wouldn’t be so enamoured with your current administration that is looking to the 1930s for its administrative policy and looking to the 1880s for its industrial policy), and remember the beginnings of trade. It was horse and carriage (well, mule-and-wagon or donkey-and-wagon) until we got the first cargo ship, which could move mass cargo by sea. Great for port cities, not so great for inland cities. Then the train was invented, and that revolutionized transport (and then travel). Locomotives quickly became more and more powerful, standardized tracks allowed them to run coast to coast, and up to 200 cars of cargo and people could be carried at once, especially if multiple locomotives are used. TWO HUNDRED RAIL CARS.

A flatbed rail car can be up to 89′ in length and 10′ wide.

A standard cargo container, used on ships, is 20′ x ‘8 or 40′ x 8’. A properly engineered flatbed rail car can hold two long or four short containers.

A typical long haul transport truck today is 53′ x 8’6″ (x 13’6″ high). No reason the trailer can’t be replaced with a specially designed 42′ x 8’6″ flatbed that could lock and load a standard 40′ container or that automated systems to lock and unlock couldn’t be designed to easily allow movement between both ships and rail cars AND between both rail cars and trucks. This would considerably shorten the distance that 400 containers (200 flatbeds x 2 containers each) would need to be transported across American roads, and significantly free up the availability of 400 drivers per train (and corresponding lane).

An average long-haul route in the US is 500 miles+! (With many routes up to 800 miles, or more).

An average short-haul route in the US is closer to 150 miles.

Long haul trucking could be reduced by 2/3 if rail was used more and all routes were short haul! Considering long-haul trucking accounts for about 200 Billion miles a year in the US, that’s 120 Billion miles that can be freed up, which greatly reduces the driver need! (If a driver drove 60 miles/hour for 50 weeks a year, that’s 120K miles.) In fact, it reduces the need by almost 100K drivers! It might not solve the entire problem, but it would be a huge dent!

It’s time we start looking back as well as forward if we want to solve the problems of today!

The reality is that over 500 BILLION miles of annual trucking is just too much! Almost 73% of freight by weight should NOT be moving by inefficient truck transport! Trucking.org has some good, and scary, statistics.

This post first appeared in a slightly abbreviated form on LinkedIn.

There are MANY reasons you are NOT ready for AI!

A few weeks ago, we told you that if you think you’re ready for AI, you’re not ready for AI because, even though the vast majority of you are chasing AI, only a minority of you are ready to even investigate it. And we mean investigate, not use. That depends on whether or not there are any relevant AI solutions for you needs — and despite the repeated BS claims by the big AI vendors, there may not yet be any!

And it’s not just because you haven’t

  • admitted you’re only chasing AI because of FOMO and FUD
  • assessed where you are
  • realized you are generations of tech behind
  • determined you just don’t have the right resources

But it goes beyond that.

In order to have any hope of succeeding with AI:

You need great data and great Master Data Management
… but you don’t even know where your data is! You have no governance policies, no management processes to ensure data is kept up to date (or even backed up unless you have already suffered a data loss and determined losing that specific data would be disastrous), and no clue about what that entails. And even if you realize that you need (master) data management, you won’t get the C-Suite to sign off on it, even if you call it E-MDMA and tell them they’re getting free samples!
You need a good IT infrastructure, with context-based integration and workflow capability
… but you have no central strategy for data integration, system orchestration, or enterprise workflows, and your IT infrastructure is whatever cloud your ERP runs on. AI, especially Gen-AI, requires massive data and massive compute and, guess what, that requires massively powerful, solid, infrastructure — and yours is probably held together with spit, glue, and duct tape!
You need an in-depth understanding of not only the problem you want to solve, but what AI algorithm will actually work reliably and with measurable confidence
… but guess what? In order to properly evaluate AI, you need an advanced understanding of the technology, which usually requires an advanced, graduate level, understanding of the underlying mathematics as well as deep understanding of the problem and how to mathematically model it.
You need a strong technical quotient (TQ) to implement, train, and verify those AI algorithms
… and that’s more than just a single expert who can evaluate, but a strong bench of architects and developers to make it work — you can’t rely solely on the vendor as they can go away, their bench can leave, or they can get pressured by their investors to just sell, sell, sell (and pretend you don’t exist once they get the cheque) and that leaves you to your own skillsets.
You need domain experts on hand to verify the results
… and this goes double for critical results. If you are using an augmented intelligence to help with sourcing, market analysis, strategy recommendations, etc. you can’t let an agentic system execute on a computation without verifying it. No system ever has all the data, no system ever knows all of the options, and no system has the soft information (and how you might be able to work a sales rep to your advantage). And if someone messed up the data, considering just one wrong number can entirely throw off a hundred thousand variable model, you’re in deep doo-doo if the system executes an order without your verification.
You need to redesign your processes to optimally take advantage of AI
… because your processes come from the time before office machines existed, so obviously they weren’t designed for modern technology. And while traditional workflow / RPA can easily automate what you have (even though it shouldn’t), since AI requires good data, good structure, properly designed models, etc. — it’s not going to work with whatever Guilded Age process you’re using now.

And so on. The reality is, despite what all the big vendors, big consultancies, and big analyst firms tell you — you’re just not ready for AI. (And definitely NOT ready for big bang projects that will end in big busts!) It’s just the latest silicon snake oil panecea — like all purpose predictive analytics, the fluffy magic cloud, SaaS, and the World Wide Web and every other panacea that has come before. (Just remember the last time silicon snake oil was hyped this much, it resulted in the dot com bust!)

Dangerous Procurement Predictions Part IV

As per our first three posts, if you read my predictions post, you know SI hates predictions posts. It fully despises them because the vast majority of these posts are pure optimistic fantasy and help no one. Why are the posts like this? Because no one wants to hear the sobering reality off of the bat in the new year and the influencers care more about clicks than actually helping you.

But the predictions are not only bad, they’re dangerous if you believe them. So we are continuing to lay bare the reality of the situation to make sure you understand that this year isn’t much different than last year, no miracles are coming, and only hard work and the application of your human intelligence are going to get you anywhere. Today we tackle the next set, and we hope we’re at the end of the series, but if we stumble across more bad predictions, we’ll have to do a Part V. But we hope not!

11. Negotiation gets productized.

Here’s the thing, in a few niche industries like electronics, we have a few niche players like Levadata that bundle “should-cost” + playbooks + concession sequencing for experienced buyers to help them leverage the state of the market for the best results possible. But they’re hardly used relative to the total electronic market size, as they are used mainly by component buyers / manufacturers, not consumers of such tech (to understand the manufacturer’s margins).

Similar offerings don’t exist across most industries. And even if they did, most buyers are not sophisticated enough to do this. Most struggle with a multi-round RFX, yet alone detailed should-cost/target cost models, negotiation playbooks (which have to cover all standard market conditions and unique situations), and the concept of BATNA, especially relative to offers and counter-offers in a structured concession sequence.

Without these domain relevant niche offerings and career negotiations trained in deep tech, which are both few and far between, this is not going to happen. And Artificial Idiocy certainly isn’t going to fill the gap!

12. AI As a “Governance” Engine.

The claim: When you design them well, agents encode judgment, compliance and brand values into every transaction. Uhm, no! At least not if they are Gen-AI agents that can’t judge (as they can’t even reason), may or may not execute compliant with regulations, and will happily screw a supplier (by refusing to pay an invoice) or customer (by refusing to honour a claim) if it thinks that’s what it needs to do to make you happy or stay turned on (because it was told to find savings of 500K and it’s calculations determine that paying certain invoices or honouring certain claims will not allow that savings goal to be met, if it was even possible when the AI told you it was as it may have arbitrarily multiplied a calculation by -1 just to make the math work).

Governance, by definition, requires the act of governing. And governing, by definition, requires the wisdom as well as the authority to conduct the affairs of the organization. And only truly intelligent beings (i.e. HUMANS) can acquire wisdom over time.

13. There will be no more “X” employees because AI will replace them all!

First of all, how many times do we have to repeat that there are NO AI Employees, you shouldn’t believe the degrading, demeaning, and, frankly, dehumanizing claims, and that you definitely DO NOT want Agentic Buying through fake AI Employees. Secondly, it can’t even do the basic tasks that even the dumbest drunken plagiarist intern can do on a daily basis. But let’s not digress too far before giving you the major examples.

Claim #1: Contract Administrator / Staff Attorney

THE PROPHET has been trying to Kill ALL the Lawyers for quite some time now, and it seems he’s not alone.

But here’s the thing. While AI systems are pretty good (and as good as the drunken plagiarist interns) at spotting grammar errors, redlining against standard clauses, pointing out missing clauses in most organizational contracts, etc., they aren’t good at everything. They can’t identify unaddressed risks without being told what those risks are, they can’t judge the full extent of liability without understanding what those liabilities could be, and they can’t judge the supply geo-political and supply chain risks without broader context.

Plus, they can’t always back up their suggestions; often make up case law, case decisions, and authors; and can’t always judge the requirements of potentially relevant regulations. And we’ve seen many times what happens when even trained lawyers use AI — they get lazy, fall for the slop, get reprimanded and fined by judges tired of the laziness (with a recent example happening in November in Mata v. Avianca, Inc). The previous link also lists three other notable cases where lawyers (and their firms) were fined and sanctioned, but, by now, there are dozens!

But hey, go ahead and replace your lawyer, write bad contracts, make decisions on fake case law, and risk your entire business if you want to. (If you want to, it’s probably safe to go ahead and get rid of the intern who does the redlining and the clerk that does the filing, the AI is probably just as good at that, but do not ever, ever replace a real qualified lawyer with a piece of sh!t “AI”.)

Claim #2: Spend Analyst

Sure you can buy auto-classification that might get to 95%, auto-cubing that can build any cube you can imagine, auto-analytics that can run the entire slate of standard analytics and compute past, current, and projected costs against past current, and projected market data based upon current buying patterns and suggest items, categories, and/or suppliers to (re) source, switch from/to, and possibly (re)shape demand.

But this doesn’t mean that it’s the right items or categories to chase, the right suppliers to use, or even the right area to focus your efforts. It’s based on math, and an assumption of consistent, stable, market conditions, but those don’t exist anymore. If you’re not also considering geo-politics, natural disaster risk, uncertain logistics when the panama canal reaches historic lows for much of the year, terrorists block the Red Sea, and unpredictable weather make sailing around the capes more dangerous than other, and sourcing for resiliency and not just cost, your “spend” analytics are useless. You need an analyst with a good understanding of economics (and access to an economist), geo politics (and access to local experts), and resiliency, not just total cost of ownership buying. (Now, the junior data pushers are probably all dead and gone, but not the real experts!)

Claim #3: Sourcing Event Manager

Now, transactional buyers are gonna get replaced by autonomous systems that use next generation (advanced) robotic process automation enhances with machine learning in Agentic systems, because ordering off of contracts, ordering from catalogs, and doing low-cost non-strategic buys through quick-quote RFPs doesn’t take any brainpower whatsoever (making it perfect for AI that has none).

But strategic sourcing requires more than just buying off of contracts, ordering from catalogs, and issuing quick-quote RFPs! It requires defining key criteria (that go beyond what engineering, marketing, or maintenance provides), identifying validated suppliers (or identifying suppliers that can be easily validated), holistically analyzing the market conditions, determining the best event type, determining the negotiation strategy, etc. The tools might be able to help with initial supplier identification, collecting numerical (commodity) market data, letting you know what event types were run in the past, compiling fact-based playbooks, and, of course, automating each extent of the process, but they can’t do real strategic sourcing that requires real human intelligence. And with today’s geo-political uncertainty, that human intelligence is needed more than ever which means that expert sourcing professionals are needed more than ever. (But dumb buyers will join the dodos.)

There are more ridiculous claims, but you get the point. Skilled jobs are not going away. (But bit pushers are.)

14. New standards for Ethical and Sustainable Supply Chains.

In some countries, current standards aren’t even being met. Good luck getting new standards introduced, since there aren’t a lot of global internationals (with those headquartered in the US in particular) that want even more rigour, especially if it will cost money! As long as laws are being minimally met, or reasonably-sized “facilitation payments” can make problems go away, this is not a priority, especially if going beyond would cost more money!

15. The “AI Singularity” is coming faster than we can process.

It’s not, because the models can’t get bigger, there is no more data, and no one has yet come up with a model that has any hope of even getting close to the actual intelligence of a pond snail.

Plus, if it ever did happen, considering a “singularity” is actually a black hole, it would rapidly consume (i.e. destroy) the Earth, and we wouldn’t have to worry about it. This is just more nonsense from the A.S.S.H.O.L.E.

C-Suite Only Has Budget for AI? Then Lie its AI and Buy Solutions That Work!

You need modern tech more than ever, but with unemployment on the rise (and people buying less), recessionary fears, tariff wars, etc., the C-Suite doesn’t want to give you the increased budget you need to buy the tech you desperately need to be more productive.

On the flip-side, they are finding millions of dollars for “AI” (in the hopes the lies are true and they can replace you, even though any attempts to do so will result in massive negative repercussions) despite the fact you’re not ready for AI and most AI-first vendor solutions SUCK, and trying to bring in in big consultancies (who will propose multi-year big-bang projects that, like all big bang projects that came before, will result in big busts and possibly create some of the biggest supply chain disasters of all time). You know the tech doesn’t work. You saw the 6% success rate from the recent McKinsey study and the 5% from the recent MIT study (both in late 2025). That’s a 94% failure rate, which is even worse than the general tech failure rate of 88% (as per a Bain 2024 study)!

You know you need modern tech, and you know, for the vast majority of those needs, AI ain’t it. Especially since you know that there is no such thing as Artificial Intelligence. (Artificial Idiocy for sure — it’s called Gen-AI — but not Artificial Intelligence. The best you can get is Augmented Intelligence, but that’s always narrowly focussed and quite rare due to too much research, and promotion, of Gen-AI LLMs that will never work [as the models are foundationaly flawed and there is no more data to train them on].)

So what you do?

Frankly, you lie and say its AI!

The fact of the matter is, if the C-Suite is insisting on AI, it’s because they don’t actually know what AI is. (This should be abundantly clear by the fact we have a lot of vendors, and consultancies, claiming AI Employees, and we all know that’s pure bullcr@p!) And the reason they are insisting on it is because everyone else is lying to their face. The Big (and small) “AI” Tech Vendors. (Big) Analyst firms. Big Consultancies. The Media publishing the fake claims and fake stories (and presenting what are bullcr@p-filled advertorials as vetted and verified case studies). Influencers spreading the hype for their own profit.

And since this C-Suite has a lower TQ (Technological Quotient) than an average elementary school child (who can probably use your phone and tablet better than you can), it’s not like they have any real clue what AI is.

And if the C-Suite wants “proof” that the vendor you select has AI, ask the C-Suite what that means to them. Nine times out of ten it is a chatbot interface. If the vendor you select hasn’t done so already, just have them hook in Chat-GPT through an API, let the execs play for 5 minutes, get the C-Suite approval, and then have the vendor disable the interface before delivery (in platforms where it truly is useless, or limit the chatbot interface to help queries where there is very little downside to it screwing up).

When Chat-GPT first became the rage in Procurement, C-Suites insisted on “AI Guided Buying” even though, with a well designed federated catalog (that support standard service forms) — that supported contracts, preferred vendors, (learned) business rules, and budgets — it was five to ten times faster to use the integrated search bar and filters (as per our twelfth entry of our 2025 Myth-busting series where we illustrated just how dumb the Gormless AI could be). After losing out on a few deals due to this lack of functionality (even though they were the buyer’s choice), a fed-up vendor built a chat-bot led guided buying offering on standard LLM libraries. They can turn it on, demo it for the pointy-haired bosses, and turn it off again.

This is critical when the real value of

  • analytics is exploration
  • sourcing is optimization (not error-prone calculations by an LLM that might erroneously multiply a number by -1 because it’s interpretation of the request is to satisfy an arbitrary savings number anyway it can)
  • supplier management is potential issue detection and human review and remediation
  • procurement is honouring contracts, using vetted suppliers, and following rules designed to prevent risk
  • etc.

So take advantage of the fact that they don’t have a clue what (real) AI is and tell them whatever tech you need to solve your problem, regardless of how much AI it does, or doesn’t, have is the latest and greatest AI if that’s what it takes to get the tech you need to solve your problem.

I know you don’t want to lie, but the reality is that is now what you have to do to keep your job, because if you don’t get the tech you need, you’ll fail and be made redundant. And since everyone else is lying too, chances are someone is already saying the tech is AI and you can just point to them (and blame them for the lie).

End of the day, it’s whatever makes you the most productive. That might be a new AI solution, might be a classic ML or NN solution, or it might be two-decades old rule-based automation where you can encode a few rules and have the solution do 95% of your work on auto-pilot without any worry of it ever screwing up (and costing someone their job).

Remember, you’re hired to get things done, not listen to bullsh!t that comes straight from the A.S.S.H.O.L.E.. Don’t get blinded by the hype!