Category Archives: Technology

AI is Not AI … And AI-Related Tech is Not Created Equal

Joël Collin-Demers recently made a post that correctly stated that real “AI solutions” tell you exactly what they do, in which sequence, and [help you] understand how it solves your exact problem. the doctor totally agrees. Otherwise, they are using buzzwords and trying to cash in on the hype to sell you old-school automation at best, or third party (Gen-AI LLM) wrappers at worst, and don’t have any real AI.

He covered ten different types of technology and attempted to capture the positives, the negatives, and the best uses therefore in source-to-pay. For a relative non-techie (compared to the doctor with a PhD, degrees in CS and Mathematics, and actual experience implementing everything but BS LLMs from scratch), he got a lot right. But he got a few things wrong. As a result, there was a need to correct him (in this post) and ensure the corrections have as much permanence as the original post.

We do recommend you read his original post, but for each tech addressed, integrate the correction below.

RPA – does not “break” when processes change; it breaks when you feed it bad data; when you change processes, it simply loses its usefulness until you change it to match the process — with a good RPA system, that shouldn’t be hard

Machine Learning – requiring clean data is NOT a bad thing; it ensures the algorithm “learns” the patterns you need it to learn to use it effectively

Natural Language Processing – doesn’t struggle with jargon, just context — you define the dictionaries, the grammar, the language — it’s accuracy boils down to that; if you are feeding in documents that use the same words/phrases in multiple contexts, it will always struggle with that to a point

Predictive Analytics – in lay terms, classical predictive analytics is essentially just multi-dimensional curve fitting based on the data available — when something has not been modelled, there is nothing the algorithm can learn to fit against — but to be fair, nothing you’ve listed will succeed with unprecedented events/data

Anomaly Detection – this is based on outlier detection, and well trained outlier detection does NOT have high false positive rates (which are no higher than false negatives), and any “wrong” classifications from a business perspective simply means that the definition of a valid transactions needs to be amended to reduce the “outliers”

Computer Vision – lighting is not as much of a problem as you think as most good algorithmic interpretations will always mathematically adjust the brightness and contrast to a consistent range in pre-processing, and sometimes even greyscale; angles are a problem, because if they can’t be determined, the right transform can’t be applied to appropriately orient the image to maximize identification likelihood

Optimization Algorithms – “requires precise problem definition” is not a bad thing, it’s a good thing — if you don’t have a precise problem definition, you cannot get a precise answer with any technique; one of the best uses is product mix, not just supplier portfolio

LLMs – not “can” hallucinate false info; “will” hallucinate false into — every single time, just a question of the degree; it’s “generative” AI which literally means it makes stuff up, and how accurate what it makes up is with respect to your problem depends on how it was trained, what was asked of it, and how you ask it … very unreliable all around

Pattern Based Recommendations – the whole point is to “filter” to what you would normally buy so that you don’t have to sort through everything, that’s not a problem

Reinforcement Learning – it does not require extensive time, it requires extensive data — computers process mathematical calculations billions of time faster than we do — it’s never time anymore!

Execution Capacity has Always Been the Competitive Advantage in Supply Chain

And The Key is Still Automation, NOT AI!

A recent article over on Global Trade Magazine on Why Execution Capacity Is Becoming the Next Competitive Advantage in Supply Chain gets a number of things right.

1. Procurement and Supply Chain leaders are constantly being asked to do more with less, and, yes, this has been going on for years (and, to be precise, decades).

2. They are managing larger supplier ecosystems, responding to geopolitical disruptions, navigating inflationary pressures, adapting to shifting tariffs, and controlling costs across increasingly complex global operations. At the same time, executive teams expect procurement organizations to move faster, identify new savings opportunities, and strengthen business resilience.

3. The challenge is capacity.

Most enterprises already have capable procurement teams, well-defined sourcing strategies, and clear objectives. What they often lack is the bandwidth to execute those strategies consistently across thousands of suppliers, transactions, and commercial opportunities.

With one supply chain catastrophe after another of the man-made and natural variety (port strikes, border closings, tariffs, wars, strait and canal closings, factory fires, droughts, wild fires, volcanic eruptions, earthquakes, tsunamis, mine collapses, etc.), the constant uncertainty in your supply chain and underlying costs, and supply lines disappearing without warning, execution gaps are becoming increasingly visible, disruptive, and costly.

In order to manage these turbulent times, your organization needs alternate suppliers, alternate supply lanes, the ability to re-allocate orders daily, the ability to re-route shipments in real-time, and the ability to optimize your suppliers, supply lanes, order allocations, and shipment routings. And, most importantly, the ability to identify, and manage, these (alternate) suppliers, supply lanes, re-allocations, and re-routings across all products that must be sourced globally! In other words, execution capability across the Procurement and Supply Chain organizations.

But, as the article notes, in an average organization, only so many new suppliers can be identified, existing supplier relationships can be optimized, products can be strategically sourced, orders can be re-allocated, and shipments re-directed.

This is because, while most organizations have invested heavy in classic analytics, supplier intelligence, sourcing platforms, and risk management tools, they have simply invested in visibility, not execution!

According to the article, the answer is to go from “AI Assistance to AI Execution”. But that’s not the answer. It’s not “AI Assistance to AI Execution”. It’s “Tech Assistance to Tech Execution”, whatever form that tech may be … and for the most part, it’s classic automation, which, we’re sad to say, has existed for over a decade, and been largely ignored for that time.

Let’s take each of these requirements one by one:

Supplier Discovery: when the organization needs to source, or re-source, a product, the tool automatically searches the supplier network for all suppliers that supply a similar product and then weights them on key dimensions of product similarity and organizational supplier scoring criteria for suppliers in that category (based on information on the supplier in the network)

Supplier Optimization: for a product/category, automatically run analyses that identify the right mix of current and potentially new suppliers based upon a combined ability to supply the organization’s demand with enough “slush” to allow for a supplier becoming unavailable due to supplier issues, supply chain issues, or other issues without adding unnecessary bloat to the supply base. (Considering that organizations typically spend 80% or more with 20% of suppliers, most organizations have too many suppliers but not enough for key products or materials.) This mix will be automatically optimized with the right automation solutions.

Order (Re) Allocation: re-run forecasts weekly/daily, re-allocate orders based on stock-levels, probabilistic forecast predictions, current and expected lead-times, expected supplier/lane availability, contractual commitments, etc. and choose a balanced solution that will satisfy all the probable outcomes (using optimization, not random AI predictions)

Real-time Re-routing: for every multi-modal lane, re-routings can happen at every waypoint (where modes shift, cross-docking at warehouses/FTZs is utilized, or where stops occur); re-run the models based on supply chain updates daily and if carriers/routes for segments are expected to become unavailable, costs become too high, or delivery times would stretch out too long (or could be stretched out to lower costs), possibly issue re-routing orders

Required Data: Automation can automatically pull/push data on a daily/real-time basis

When you consider that modern AI falls into Gen-AI which is literally “make stuff up”, you can’t depend on it for critical supply chains where one mistake can be catastrophic. But, fortunately, there are systems out there that do all of the above reliably on classic RPA, optimization, and analytics. (And have been for about a decade.) Plus, with the recent SaaS price compression as a result of the AI Hype wave, it’s all very affordable.

So if you want to succeed, get these systems. They’ll allow you to manage all suppliers, all items, and all lanes. You’ll be able to execute on your strategy, provided you can come up with a strategy that is adaptive enough in today’s global economy.

If You’re Spending 250K Annually Per Engineer On AI …

Then not only are you contributing to planetary destruction (through the generation of between 1.32 tons (high end models, 1 joule per token) and 84 tons (low end models, 2 joules per token) of CO2 to power those data centres, which is about 0.2 to 12.7 times the average individual carbon footprint, with an expectation of 7 to 11 tons (Source), and the utilization of 300,000 gallons to 5,000,000 gallons of water a day to keep those servers cool, or a town’s worth of water every day!

BUT YOU ARE NEEDLESSLY WASTING 400K+ A YEAR

1. Less than 20% of AI generated code survives unscathed in a commercial enterprise software product once senior developers weed out all the security errors, boundary condition errors, and generated code that doesn’t even solve the problem. So, that’s 200K of 250K down the drain as only 20% of output is usable.

2. Having to fix AI generated slop will consume 80% of a good senior developer’s time — a developer you should also be paying 250K a year.

End result, you’ll losing 200K + 200K per developer you force AI coding tools upon!

But hey, it’s your money. If you want to p!ss it away so NVIDEA’s CEO can get richer selling more CPUs we don’t need, that’ up to you!

The linked article contains some metrics, but here are a few others.

  • token prices vary widely, from an average of around 50c/M tokens on the smallest, cheaper models to $75/M tokens (or higher) for higher end “workhorse” models
  • energy processing requirements per token are estimated to be between 1 joule and 2 joules
  • you can buy 14.3 Trillion tokens at the median of around $17.5/M tokens (and 35 times that at the lower end)
  • processing 14.3 T tokens will take about 4000 kwH @ 1 joule/token
  • on an average NA grid, expect to produce 500 to 600 g of Co2 per kWh (since most of our grids are still dirty)

Exact Purchasing Helps You Define Your Tech Needs

In our last post we illuminated how Busch-Lamoureux Exact Purchasing required Category Intelligence (not just the Category Management most Procurement organizations aren’t yet doing) and, thus, ups your Procurement game in more ways than one. Many more ways than one, actually.

Only with Exact Purchasing can you figure out what you actually need from your Procurement technology, If you go back to our piece on assisted solution selection is a seven stair methodology, step 1 is understanding your needs. By breaking your procurement needs into categories you can specify to a high degree of detail, you get an understanding of what you really need to do in your Procurement organization.

In the seven stair methodology, step 2 is the holistic solution requirement — and this is what is embodied in Exact Purchasing! With exact purchasing, you are holistically evaluating a category from all key perspectives — complexity, risk, and organizational impact — and creating sourcing, procurement, supplier, and supply management plans that balance the requirements from a holistic requirement.

Step 3 is organizational maturity, and here’s what most Procurement organizations miss — the lack of a proper, formal, category management strategy that allows them to start on the journey to Procurement excellence through better processes, risk management, and intelligence is what holds them back. Exact Purchasing gives them a foundation to not only figure out where they are on their Procurement journey but where they need to go and what process improvements might get them there.

Step 4 is vendor pool selection, and here’s where exact purchasing really starts to help as it helps you identify what the tech has to do, which helps you (possibly with help from an expert analyst or consultant) identify what type of Procurement tech you really need, and then you can use an independent analyst or consultant (who doesn’t have to sycophantically cater to the bejeweled emerald software partner in order to maintain that bejeweled emerald status that sees a lot of integration work thrown his way as long as he maintains it) to identify the vendors most likely to be a great fit for you.

Step 5 is the vendor assessment process, which is itself a 7 step process — and Exact Purchasing helps you out end-to-end here.

Step 5a is RFI creation. With Exact Purchasing, you know what the critical functionality is, you can easily specify what it is, and then quickly eliminate any vendor that can’t meet 100% of the critical must-have for your organization before wasting any significant time on them.

Step 5b is collaborative RFI review. Once you’ve eliminated those that you’re certain won’t fit, if too many vendors survive the cut, the team is educated on both what is needed and what will make their lives easier and can holistically assess initial responses to narrow down to the providers that go beyond the basics in ways that might be helpful.

Step 5c is the qualifying demo. You can create a script that not only covers all the essentials, but should haves that will help illuminate where the key strengths and weaknesses are likely to be both in the given vendor’s application but the vendor pool over all and get the insight you need to ensure that you’re both on the right path and that the vendors you select for the RFP will be worth the next stage review.

Step 5d is the RFP creation. From here you can elaborate all the should have and nice to have functionality, double down on your key pain points that you would like solved (potentially in innovative ways), note what intelligence is critical, identify where you’d like services and support, and identify any must-have organizational requirements beyond Procurement that would be a deal-breaker. You’re able to focus on the what, instead of the typical 500 point feature list where half the features you might never use.

Step 5e is the RFP review, where again the team has the understanding on what to look for and can ensure that any vendor who wouldn’t make the cut doesn’t get invited to the full demo stage.

Step 6 is the full demo where each vendor provides a two-part demo against the basic deep dive should have script the RFP was based on and detailed requests based upon claims they made in the RFP against nice-to-haves, uniqueness, process improvements that will save you time and money beyond what peers can do, etc.

Step 7 is the decision, and you can make that against what the vendors offer relative to your category needs and organizational goals, not just a feature list you don’t understand (but copied from a Free RFP anyway because you needed to look competent).

In other words, Exact Purchasing gives you the understanding you need to go to market, and the understanding that not every solution may be appropriate for every category — and that’s okay. Sometimes two or three targeted BoB solutions are still less than half of the cost of a mega-suite that still only solves half of your problems.

The One Big Benefit Of NOT Going AI …

You don’t have to worry about your AI vendor going toes-up when power costs go through the roof and your AI vendor can no longer charge pennies for compute when its costs rapidly become dollars and it can’t pass them on due to contractual commitments to existing clients (or to new clients who won’t pay dollars for computations that might return hallucinations).

The new generation of AI tech — Gen-AI LLMs / AGI — requires way more compute power than the last generation, 100 to 10000 times more on average, for most requests. Grids are stretched and beginning to break. We’re at the point where only nuclear can power the data centre needed for a modern Gen-AI/AGI offering. And, as per Koray Köse’s recent article on AI leadership is about who controls the power, U.S. nuclear plants operated at 92.3% capacity last year. OUCH!

THERE IS NO ENERGY LEFT!

You can’t build a new nuclear plant overnight — if you can even build one at all anymore! Last year, DOGE’s Firing Fiasco at the NNSA stretched an already stretched organization even more. Many returned to work, but not all, but budget cuts likely left them without the capacity to even properly monitor existing aging nuclear infrastructure, yet alone approve more plants.

And it’s not even clear how much know-how is left in the US to build new plants. The Vogtle Units 3 and 4 in Georgia were the first units built from scratch in over three decades. The experience and expertise isn’t there to safely build these plants en-masse.

And the last thing the US wants to risk is another meltdown. Three Mile Island wasn’t a Chernobyl, but all it takes is a rushed private sector job with a lack of proper oversight and testing and one small mistake to trigger the next meltdown on US soil.

In other words, the power isn’t there for more AI.

So those organizations that can do without modern AI, that can use classic solutions with fit-for-purpose last generation AI that requires a fraction of the power and can run on already strained, non-nuclear, grids will be the big winners when the power squeeze hits and the Big AI players start dropping like flies.