On noise, NPCs, chatbots, and supply chain
The state of supply chain communication is dismal. By way of anecdotal evidence, for the last couple of months, my daily morning routine consists of attempting to find one post – just one – that qualifies as worthy of interest to an audience of supply chain minded people – in order to have it shared on news.lokad.com. This seemingly simple exercise proves exceedingly difficult. Most days, I do not succeed.
Yet, there is no shortage of supply chain communication. On the contrary, consultants, software vendors (including yours truly) and academics1 mass produce content on a daily basis. Every day, over various social networks, I scan a few dozen examples of this communication 2 – sometimes getting close to a hundred if time allows – and here is my verdict: an unmitigated torrent of white noise 3. This field has been overrun by the “content creator” mindset that I would summarize as the art of turning platitudes into themed attention-grabbing pieces.
For years, I have been intellectually lumping individuals and organizations playing the white noise content creation game under a single unifying banner: the horde of NPCs (non-player characters). Devoid of agency, they go with the flow, relentlessly spinning vapid pieces, themed with the narrative of the day4. They go through the motion of whatever format is expected of them, following whatever themes seem to be fashionable. The term “NPC” isn’t nice but generating white noise is worse, a lot worse.
Yet, reality is stranger than fiction. Decades in the making, chatbots 5 are finally capable of producing written pieces that are near-indistinguishable from what humans produce. Some historians may declare 2022 as the year that some AIs started passing the Turing Test, but I won’t 6. It’s the other way around: in 2022, most content creators stopped passing the Turing Test. However, this new wave of chatbots has proved that mainstream content creation requires neither sentience nor intelligence.
The NPC hypothesis was correct all along.
As chatbots dramatically lower the cost of “mainstream” content creation, corporate content creation will soon get turned up to eleven. Furthermore, as supply chain is an elusive subject at the crossroad of many corporate struggles, it will most certainly prove to be one of the fertile grounds for the automated mass production of platitudes. Like winter, Captain Obvious 2.0 7 is coming, and this time, he is in command of a chatbot army.
Yet, high quality communication matters for the continuation of any company - let alone its improvement. This take is even more acute in the case of supply chain. Indeed, supply chain experiments – which could be used to validate if ideas, technologies, or processes effectively work – are difficult, costly, and usually impractical to carry out. Bad communication, conversely, allows terrible ideas, technologies, or processes to survive much longer8 than they should.
Thus, my proposition is the following: as the rise of chatbots is now inevitable, companies that fail to develop intolerance to white noise will gradually regress on all fronts. Divisions, like supply chain, where real-world feedback is delayed and diffuse, will regress faster.
Intolerance is stronger than indifference. Indifference means moving on without further interactions. Intolerance means banning further interactions. Indifference to white noise is not sufficient: whenever you interact with a piece produced by a chatbot, even if it’s to immediately discard it, whoever controls the chatbot wins. Indifference only delays defeat. This sort of battle can only be won through systematic retaliation.
Retaliation, in business, is not nearly as dire as it sounds. All it takes, to retaliate, is to stop sending money to the offending party. For example, if vendor XYZ is spinning white noise on social network ABC, then don’t renew existing contracts and don’t sign new ones either.
Don’t feed the troll.
Annex: ChatGPT discusses supply chain
The discussion below was produced with the help of a third-party chatbot on December 16th, 2022. Any similarity to the content that can be found on your website is purely accidental. I solemnly swear that I have not cut-and-paste any material(s) from your website. While similarities are extensive, some might even say troubling, they are nonetheless entirely fortuitous.
Best regards, Joannes Vermorel
Joannes Vermorel: Explain what my company needs to improve its supply chain.
ChatGPT: There are a number of factors that can impact the efficiency and effectiveness of a company’s supply chain. Some things that your company may want to consider improving include:
- Visibility: Having real-time visibility into the location and status of inventory, orders, and shipments can help your company make more informed decisions and respond to changes in demand more quickly.
- Planning and forecasting: Accurate forecasting of demand and production can help your company avoid overstocking or running out of inventory, which can lead to lost sales or costly expedited shipping.
- Collaboration: Collaborating with suppliers and customers can help your company identify and address bottlenecks and other issues in the supply chain, and can also help you negotiate better terms and prices.
- Transportation and logistics: Optimizing the transportation and logistics of your supply chain can help reduce costs and improve efficiency. This may involve choosing the most cost-effective modes of transportation, using technology to track shipments and optimize routes, and implementing just-in-time delivery.
- Inventory management: Proper inventory management can help your company reduce waste, improve customer satisfaction, and increase profitability. This may involve implementing techniques such as just-in-time inventory, cross-docking, and consignment inventory.
- Quality control: Ensuring that the products and materials your company uses are of high quality can help reduce defects, returns, and customer complaints.
- Sustainability: Implementing sustainable practices in your supply chain can help reduce costs, improve your company’s reputation, and support long-term business growth.
Joannes Vermorel: Explain what my company needs to make its truck shipment arrive on time while shipping from Paris to New York?
ChatGPT: There are several things your company can do to ensure that its truck shipments arrive on time when shipping from Paris to New York:
- Choose a reliable carrier: It’s important to select a carrier that has a track record of on-time delivery and that can provide the necessary documentation to clear customs.
- Plan ahead: Allow plenty of time for shipping to account for any potential delays. This includes allowing extra time for customs clearance and other unforeseen circumstances.
- Use tracking and monitoring tools: Using tools such as GPS tracking can help you monitor the location and progress of your shipments in real-time, which can help you identify and address any potential delays.
- Manage the paperwork: Make sure all necessary documents, including customs paperwork, are in order before shipping. This can help reduce delays at customs and ensure a smooth transit process.
- Communicate with your carrier: Keep in touch with your carrier and stay updated on the status of your shipment. If there are any issues or potential delays, they may be able to work with you to find a solution.
- Consider alternative routes: Depending on the specific needs of your shipment, it may be more efficient to use alternative routes or modes of transportation. For example, shipping by air may be faster but more expensive, while shipping by sea may be slower but more cost-effective.
Joannes Vermorel: Explain how real time GPS tracking helps trucks to avoid traffic jams in the middle of the Atlantic.
ChatGPT: Real-time GPS tracking can help trucks avoid traffic jams in the middle of the Atlantic by providing up-to-date information on traffic conditions and potential delays. This information can be used to route trucks around congested areas and onto alternative routes that may be faster or more efficient.
For example, if a truck is equipped with a GPS tracking device and there is a traffic jam on its planned route, the tracking system can alert the driver to the delay and suggest an alternative route to avoid the congestion. This can help the truck stay on schedule and arrive at its destination on time.
In addition to helping trucks avoid traffic jams, real-time GPS tracking can also provide valuable information on the location and status of shipments, which can be useful for logistics planning and customer communication.
Joannes Vermorel: Can my company recycle its demand forecasts in order to become more sustainable?
ChatGPT: Recycling demand forecasts can help your company become more sustainable in a number of ways. Here are a few examples:
- Reduce waste: Accurate demand forecasting can help your company avoid overproduction and overstocking, which can lead to waste and excess inventory. By recycling forecasts, you can ensure that your production levels are closely aligned with actual demand, which can help reduce waste and save resources.
- Improve efficiency: Forecasting demand can help your company optimize its production and supply chain processes, which can lead to more efficient use of resources and lower costs.
- Enhance sustainability efforts: By using demand forecasting to optimize production and reduce waste, your company can also reduce its carbon footprint and other environmental impacts. This can help you meet sustainability goals and improve your company’s reputation.
- Foster customer loyalty: Customers are increasingly interested in sustainable products and practices, and using demand forecasting to support sustainability efforts can help your company attract and retain environmentally-conscious customers.
- It’s worth noting that recycling demand forecasts is just one way that your company can become more sustainable. There are many other steps you can take to reduce your environmental impact, such as using eco-friendly materials, implementing energy-efficient processes, and reducing waste in your supply chain.
Joannes Vermorel: In order to end unsustainable service levels, explain how companies should end discrimination against unreliable suppliers.
ChatGPT: Discrimination against unreliable suppliers can take many forms, including refusing to do business with them, providing them with less favorable terms or conditions, or not providing them with the same level of support or resources as more reliable suppliers.
To end this type of discrimination, companies should adopt a fair and consistent approach to evaluating and working with all of their suppliers, regardless of their reliability. This may involve implementing policies and procedures for evaluating and managing suppliers, setting clear expectations for performance and communication, and providing support and resources to help suppliers improve their performance and meet the company’s needs.
It may also be helpful for companies to regularly review and assess their supplier relationships to identify any issues or challenges that may be contributing to unreliable performance, and to work with suppliers to address and resolve these issues.
Ultimately, the goal should be to create a collaborative and mutually beneficial relationship with all of the company’s suppliers, where all parties are committed to meeting the company’s needs and working together to achieve shared goals. By adopting this approach, companies can help to ensure that they are able to work with a diverse and reliable group of suppliers, which can help to improve the sustainability of their operations and increase their competitiveness in the market.
Supply chain papers with formulas and algorithms certainly look a lot more “scientific” than short self-help guides repackaged with the narrative of the day. However, spinning endless variations of existing methods and models doesn’t contribute to the progress of science. It’s an elaborate mimicry of actual science. ↩︎
News trivia, case studies, announcements, white papers, scientific papers, PhD manuscripts, essays, insights, memoirs, books, video interviews, social media threads, … ↩︎
White noise is the (somewhat pedantic and nerdy) characterization of random talk without meaningful content. In signal processing, white noise is a type of signal that doesn’t carry any information and reflects a strictly random process. ↩︎
Data mining, offshoring, cloud computing, big data, autonomous computing, IoT, diversity, machine learning, inclusivity, blockchain, artificial intelligence, sustainability, NFTs, demand sensing, digital twins, nearshoring, and resilience are to name just a few themes that have waxed and waned during the last two decades. ↩︎
ChatGPT was released by OpenAI in November, 2022. ↩︎
ChapGPT is an impressive piece of technology, but make no mistake, it is first and foremost a master bullsh*tter. It displays both an advanced mastery of the high level patterns of the English language - including a super-vast array of domain-specific patterns - and a lack of any grasp on reality, even at an incredibly basic level. ↩︎
Free markets are the ultimate filter: given enough time, they ensure the termination of bad ideas and inefficient practices through company bankruptcies. However, this selection process can be fairly slow. ↩︎