How AI Is Rewiring Logistics

How is AI shaping logistics? At first glance, it looks like the effort should pay off.

In Supply Chain and Logistics, the first big shift came with ERP systems like MFG PRO, Oracle,Net Suite Epicor, IFS , Microsoft Dyamics 365 and SAP. They gave structure to data, cleaned up duplication, reduced errors, and helped teams see how information moved through the business. The next layer organized and presented that data so managers could track assets, fuel, energy use, manpower, and admin costs. Then came scenario building: what happens if you redeploy people, shift energy loads, or change routes? Once messy, conflicting data became clean and connected, the business started to look like a continuous game of balance, decisions, and trade offs.

So if the data is now clean and visible, where does AI fit?

First, AI sits on top of ERP. It does not replace it. Most logistics companies had a set of software - some even as basic as excel for many functions! - and AI is able to work on and integrate all of these to present a final coherent output.

Second, AI uncovers deeper insights, the kind teams sensed but ERPs were too rigid to show.

Third, AI adds context. ERP handled internal data well, but AI brings in industry, trade, and cultural signals.

Together, ERP and AI create a more "live" system. Internal data gains meaning because you see it against your industry position almost in real time.

With LLMs and SLMs (small language models), AI helps managers spot gaps in benchmarking and tap into broader knowledge from the internet and trained models, with care.

AI’s real promise is sharper awareness of how business levers connect. Human resources, finance, sales, operations all become easier to understand as one system.

Take telemetry. Say you have a hundred trucks on the road. Modern telemetry tells you where they are, how fast they are moving, congestion levels, and alternative routes. If routes stay stable over time, AI can predict delivery times, plan routes, and estimate fuel use from the steady stream of data coming off those trucks.

And that is only the start. Logistics generates data every second, each piece tracking a moving part of the operation. There is a lot more AI can do here, and we will explore that next.

What AI Changes in Creativity, and What It Never Will

GenAI has flipped the creative industries on their heads. Work that once demanded huge budgets, long timelines and large teams can now be done for a fraction of the cost. Speed to market drops by half. Localisation becomes effortless. At first glance it looks like agencies, copywriters, designers and technical directors should all be out of the picture.

But the truth is more balanced. Some parts of the creative process stay exactly the same. Others evolve fast.

What Stays the Same

Human imagination still sets the spark. The odd, bold, emotional leap that defines great ideas does not come from a machine. A model can remix patterns. It cannot replace the instinct that lifts a message from ordinary to unforgettable.

Screenplay and direction survive. They shift their focus from actors and technicians to model settings, prompts and software, but the role of the guiding mind remains.

Adaptation remains constant. Traditional shoots change by the minute and so do the models behind AI tools. Creative teams still need to react, refine and watch every detail.

Prompting becomes a creative craft of its own. Turning a raw idea into the right instruction takes as much care as writing a script. A sloppy prompt can wreck a project. Post production still eats hours. Rendering still takes time. The effort simply moves to new tools.

And the old complaint stays the same. “Your costs are low now, so why am I paying you more?” Buyers assume AI slashes everything. They forget that most of the creative work still happens in the thinking, not the hardware. The cost line items have shifted around, but they still remain albeit at somewhat lower levels. 

What Changes

The talent mix evolves. A new generation of techno creatives emerges. They can sketch on a page and work in Sora 2 or Gemini Veo without missing a beat. The veterans who adapt will thrive. Anyone can learn if they choose to.

Output will rise. Tech gets simpler and more powerful at once. Understanding the tools becomes essential, not optional.

Distribution shifts. What once needed cinema screens or cable TV now lands on YouTube in 16:9 or social feeds in 9:16. Every idea needs multiple versions on day one.

Production shrinks. Crews, cameras and props give way to digital assets. AI uses the same vocabulary as filmmaking, yet the “camera” is now a bundle of instructions. Smaller studios and flexible freelancer collectives will take on work once reserved for big houses.

Institutional memory spreads across people and systems. Knowledge can be stored, shared and reused in ways that were impossible when everything lived only in the heads of senior staff.

Models and film stars may no longer dictate the entire creative process, but they will protect their ground in new ways. Expect contracts to evolve fast. Rights over likeness, voice, motion capture and digital replicas will become standard clauses. Talent will safeguard how their image can be trained, generated, reused or licensed. They may have less control over production, but they will fight hard to keep control over themselves.

And yes, costs drop. At Jetmetaphy Labs we see our 70:30 rule hold again and again. AI delivers roughly 70 percent of the gain at 30 percent of the cost. It is reshaping the P&L from the ground up.

That is how the creative world is shifting. Some parts transform. Some stay rooted. And while I could be wrong, it does not feel like I am far off.

Whats the volume of AI generated ads in total digital advertising?

In all the hype about GenAI , CGI , AR, 2d/3d, how exactly does one determine what is the percentage or volume of advertising that is now AI generated? We all know its grown madly, but by how much? 

Take instagram, that behemoth of social media. Instagram isn't new to AI- its been using it steadily since 2010- 15 years ago! Remember when it could identify photos, use filters to change colour / shade, group them and so on? Then text detection from photos? And much more since those early days. Today, while it sharpens its tools in AI to provide much more stickiness to users, it also tags content appearing on its platform to identify it as AI content , as distinct from real life generated content like photos and videos as we used to know them traditionally. 

Instagram tags with "AI Info" ( changed from "Made with AI" after backlash from conventional content creators who would quite innocently do post- production changes). If human content is more than 80% its fine; more than that gets noticed as AI. While the AI - CGI generation looks very real to the eye, the metadata of every picture / video reveals the AI work done on it. 

Many AI tools embed credentials in the meta data- thats a give away that the content was made using that AI tool. Then some tools add invisible watermarks in the pixels, visible to instagram even if the metadata is stripped.  Meta itself is working with the industry to set standards. 

Then some brands actually tag it as AI generated. Thats nice. And ethical. And clear. And it still gets attention, which is the target. It also keeps the brand within safe limits, mainly from misrepresentation , penalties or deliberate out of context.

A media4u.com report quoting TAM AdEx Half Yearly Report on Digital Advertising on September 25 2025 noted that digital ads in India grew 2x in just the First Half of 2025 and that instagram handled 63% of all ad impressions. If a rising percentage of ads now use AI in some form- mildly or completely, the AI usage would be humongous already! 

We could juxtapose agency revenues / data / AI maturity/ rising AI tool sales and other digital data to determine AI volumes. My estimation is it would be well over 5x- 7x of today's usage within a year! 

But remember : when everything else fails, there's Humint or Human Intelligence.  In addition to metadata, there are ways the human eye will judge reality- that too straight a line or polished surface; that too perfect a pose /skin/ movement/ too "sweet" moments/ unnatural movements. The Human in the Loop is a totally indispensible! 

How Brands Embraced AI Ads Across Social Media

Brands moved fast to jump on AI-generated advertising. Below is a partial list of ads that appeared on Instagram between November 2024 and November 2025. There are many more across social platforms, but this gives a clear snapshot of how quickly brands adopted the tech.

Brand Name	Ad Summary	Month of Publication	Duration	Agency
Emami Dermicool	Fully AI-generated sci-fi film "Dermicool Warriors" reimagining classic jingle with futuristic AI visuals and warriors battling heat.	May 2025	30 seconds	Wondrlab, TopScout, Crushed Studios
Fevikwik	"AI Pack" interactive campaign with KwikGPT; users input objects for AI-generated mashups sticking random items together humorously.	August 2025	38 seconds	Ogilvy India
Tata Tea Premium	"Desh Ka Garv - Pradesh Ki Kala" using AI/VFX to animate regional art (Warli, Gond, Kalighat, Phulkari) celebrating India's cultural pride.	August 2025	90 seconds	Creativeland Asia
Doctors' Choice Oil	"Aamar Durga" Durga Puja campaign with AI-personalized CGI videos of actress Sandipta Sen delivering resilience messages to women.	September 2025	Variable (15-30s)	Catalyst Ad & Events
Zara Shatavari	First-of-its-kind AI cinematic ad film featuring synthetic AI model for Ayurvedic brand promotion with immersive storytelling visuals.	June 2025	Not specified	In-house AI production

source: media reports, YouTube, Instagram, my own prompt to generate list using AI engines. 

HUL's AI Initiatives Enhancing Consumer Experiences

The following is a listing of the use of AI by Hindustan Unilever in consumer facing interactions: 

source : HUL annual report 2024-25 ; my own AI prompt 

Interestingly, HUL has an AI engine for Lakme, which we will cover later. 

Inside the MD&A: The Quiet Rise of AI in FMCG

A review of the Management Discussion and Analysis sections in annual reports and filings from India’s leading FMCG companies shows a clear pattern:

• Most position AI as a support tool for analytics, measurement, automation, and consumer engagement. These efforts sit inside broader CRM and automation technology investments.
• None of the companies list AI spending as a separate line item. It is folded into overall IT and technology capex.
• Companies see these details as strategic and competitive, so they avoid disclosing explicit AI costs in public filings.

Summary of AI in MD&A : 

source: MD&A, annual reports; my own prompt for AI. 

HUL's adoption of AI across workflows

A review of annual reports, MD&A sections and analyst meetings for FY 2024–25 across the top ten FMCG companies in India shows a strong push toward AI.

Take Hindustan Unilever for example. HUL is applying AI across data flows, workflow automation and consumer-facing processes. Its FY 2024–25 annual report highlights the following initiatives:

source: HUL annual report 2024-25; compiled using my own AI prompt. 

Watch out for more detailed posts on AI adoption in FMCG in India. Should be interesting as there will be action both on workflow as well as consumer facing! 

What Is Generative AI — in Practical Business Terms

Generative AI has quickly shifted from a research concept to a core theme in boardroom discussions. The appeal is straightforward: it doesn’t just analyse data, it creates new output from it — text, images, video, code, designs and more. The underlying model learns patterns from vast datasets and uses them to produce original results on demand. This is quite distinct from Discriminative AI ( which cannot learn from datasets). 

The business interest is driven by measurable productivity. Today multiple studies and media reports indicate that generative AI can cut time to market by 50-70%. Our own work at Jetmetaphy Labs shows that marketing functions are compressing campaign development timelines and moving them in house. Software teams are reducing coding and debugging hours. Retailers are personalising customer experiences for millions without adding manpower. Even HR is automating routine documentation and internal communications.

The strategic value is beyond content creation. Enterprises deploying generative AI for decision support, customer engagement and product innovation are seeing deeper benefits. Insurance companies are using AI copilots to assist agents during client conversations. Banks are using generative systems to guide financial planning based on customer behaviour. Manufacturers are generating and evaluating product prototypes without physical mock-ups.

Adoption requires control. Effective deployment depends on data governance, privacy protections and mechanisms to reduce hallucinations. Generative AI scales best when paired with strong oversight rather than used as an uncontrolled output generator.

Conclusion : 

Generative AI is no longer an experiment. For businesses, the real opportunity lies in treating it as an enabler of productivity, innovation and faster decision-making — not as a tool for inexpensive content.

Sources : Researchgate / Economic Times / Media reports / Jetmetaphy Labs primary research