If there ever was a greater conspiracy theory than the reptiloid nature of the American government, it is AI. I mean—look no further than your LinkedIn feed. AI promises to automate everything and anything, from email drafting to writing blogs on autopilot. AI is expected to take a business’s revenue up a notch. AI is sure to amp up your sales game with human-like chatbots. From the look of it, it feels like some noname.io is all a business needs to replace its entire marketing, SMM, and content strategy departments! The big research guys, like MIT and Harvard, have already published their studies on AI-powered innovation. Frankly, the findings are conflicting with the general public’s enthusiasm for the concept.  I decided to dig deeper and actually talk to people who do stuff with their own hands. I interviewed marketers, investors, e-commerce professionals, and others to understand their take on AI. MIT Harvard The first thing I heard was Sebastian Stanley-Jones saying: “95% of AI initiatives fail [referring to the MIT study from September 2025]. AI needs to generate or save $1 trillion in the next 12-18 months to justify investment.” And that comes from the Director of EMEA at a leading digital asset marketplace! I mean—no blind guesses, but rather some educated pessimism. Sebastian Stanley-Jones 95% of AI initiatives fail . AI needs to generate or save $1 trillion in the next 12-18 months to justify investment The last thing I heard was from the organizational psychologist Vincenzo Russotto, who is adamant that humans are simply incompatible with AI. He says it takes some serious effort to transform and innovate into an AI-powered organization. No roses and sunshine, huh? Vincenzo Russotto, This is not an article to disprove AI or argue against its unstoppable transformative nature, for it is a game-changer indeed. Rather, it’s my quest to find where the truth lies. Read on to join me in my search through 11  interviews with professionals across marketing, e-commerce, consulting, and content creation. “Nope, not taking it”: Skeptics In the minds of some people I talked to, AI doesn’t live up to the hype. They spot mismatches between the promise of AI and the feedback businesses worldwide accumulate as they try to invite AI into their operations. Sebastien Stanley-Jones, who is currently the global strategist at Flippa, somewhat reflected my own doubt: AI-generated content feeds more AI-generated content, driving the human out of the automatic. Texts become soulless, blogs and posts indistinguishably the same. Further, Sebastien explains that AI is a powerful leverage, that extra brain power which performs brilliantly in coding and STEM research, like math and biology, while the commercial use of AI is controversial. In that regard, the ultimate risk is that massive job cuts, where roles are delegated to AI, can hurt the economy: “The enthusiasm is bubble-like and sentiment-driven”. Sebastien Stanley-Jones “The enthusiasm is bubble-like and sentiment-driven”. Ironically, Thaveesha Jindasa, or T.J., the founder of the Indonesian conversational AI chatbot PeakAgents, doesn’t really trust AI output quality. The busy man that he is, T.J. still writes his LinkedIn posts on his own. The rationale is simple: AI-generated content won’t rank. Compromising quality for apparent time saving is not an option. Thaveesha Jindasa “AI is a Co-Pilot”: Pragmatists AI accelerates work; there is no doubt about that. (Come on, I asked AI to help me with this article—rereading all those scripts would have taken too long!) But the secret is to let AI assist you: it may be asked to review and point out weak spots but the core action of writing and aggressive editing is on you. Matt Anderson is a strategic e-commerce consultant, and his interest in AI directly translates into the wide array of AI tools he uses or has tested at work. In his approach, Matt uses AI “as a graduate-level copilot, not brain replacement”. Matt had an Amazon agency in the past, and this proven e-commerce marketer with 18 years of experience compresses a half-day product listing task and a 5-7 day brand audit task into just 1-2 days—with AI. Despite this impressive time saving, Matt’s caution towards AI is caused by the fact that AI simply lacks context. AI works with the data it is fed, and it can’t think outside the box, the way an experienced marketer does. Hence, the limitation that keeps AI a co-pilot, rather than a co-worker. Matt Anderson as a graduate-level copilot, not brain replacement Has anyone else had such a positive experience with AI, where a human manager only stepped in at the end for checking? The content lead at Accel Club, Natali Hall, holds a hierarchical view of AI: "I'd say a solid 30% involves AI. Though, AI acts more like a junior copywriter, and it still has to be tweaked by the 'editor-in-chief' for target audience specifics and more 'human' wording”. Her next take somewhat echoes what Matt previously shared: “It's never 100% clean AI-creation because AI just doesn't know the target audience as well.” Natali Hall, I'd say a solid 30% involves AI. Though, AI acts more like a junior copywriter, and it still has to be tweaked by the 'editor-in-chief' for target audience specifics and more 'human' wording It's never 100% clean AI-creation because AI just doesn't know the target audience as well The CEO of CPGAMZ, Emad Barlas, who uses AI in e-commerce to source and analyse keywords, explains that AI “speeds up workflows” by reducing 3 days’ worth of manual work to 10 minutes with detailed prompts. In terms of personal productivity, Emad is now managing 7 clients instead of the previous 5. HR-wise, Emad shares an (alarming?) outcome: the team is now a quartet instead of a dozen it once was. Emad Barlas “speeds up workflows” In a similar vein, the fractional Head of Marketing Dana Doron calls AI “a time saver,” which nonetheless still needs human oversight. She used ChatGPT to summarize in-depth customer interviews by common sentiment: “I take the AI transcripts from those, feed them into ChatGPT, and have it summarize the information and bucket it into key findings.” Interestingly, Dana shares a situation where AI was able to make up for a tight design budget by generating far more logo designs than would have been possible for the client to commission from their designer.  Dana fed the same prompt into 5 different AI tools and, while many of the outputs were unusable, the sheer volume enabled her to find one that her client loved. As a result, a brand and package redesign project that was at a standstill was able to move forward and stay on schedule.  Was it easy? Reflecting on the process, she said, “It definitely took a lot of my time, but I learned a lot about the different models and about prompt refinement.” Dana Doron a time saver, I take the AI transcripts from those, feed them into ChatGPT, and have it summarize the information and bucket it into key findings.” It definitely took a lot of my time, but I learned a lot about the different models and about prompt refinement ‘AI Actually Works’: Innovators Now, how about AI getting into the core of a business operation? Has AI ever truly reshaped a process, giving it more efficiency and, if possible, revenue? Thankfully, I was able to get hold of Stephen Ellul. Stephen is a serial entrepreneur, and he is now an AI consultant at Breezeflo. The product he shared was a money-chasing solution that contacts clients on payment due dates. The trick is, the system learns from previous payment behavior and sends out reminders in anticipation, right when a given client tends to make a payment. While Stephen is definite that AI is revolutionizing apps and brings personalization to a new level, he makes a critical note: “Success [with AI] depends on user knowledge and prompting skills.” Just like beauty lies in the eyes of the beholder, so the true value and meaning of AI is in the hands of the user. Stephen Ellul Success depends on user knowledge and prompting skills Another avid supporter of AI is George Kapellos, and his “success story” is beyond inspirational. George is the Global Marketing and Communications Director at Nyobolt, and he is responsible for all things marketing and branding in the company. His use of AI is dictated by purely economic reasons: George explains how paying £9 for a monthly subscription to a couple of tools saves him a £300,000 agency fee for an identical scope of branding and marketing work and strategizing. The secret to a well-written, well-edited, and fine-tuned brand strategy? Rigorous prompting and mandatory human oversight: “I would never post anything or put anything in the public domain that comes from me or the organization without a human being having checked it.”. George Kapellos, I would never post anything or put anything in the public domain that comes from me or the organization without a human being having checked it Rodney Thomas, the CCO at Hi Tek Global Start Ups, uses AI as a helpful co-pilot across his work, from fine-tuning emails and messages to editing reports. Rodney’s personal caution is: “AI doesn’t necessarily have a personality,” so he always adds his own voice and his own self into what AI generates. His team’s veteran support app leans on AI heavily in the backend to “pretty much streamline the overall process”, from coding to GTM strategies and roadmap development. Strategically, AI is the team’s engine for analytical tasks, like demographic analysis, audience targeting, and quantitative insights. In contrast, the frontend unites the human touch and AI: people are at the front line and talk to users, while AI is instrumental in rapid and effective information retrieval. Rodney Thomas AI doesn’t necessarily have a personality, pretty much streamline the overall process What Does It Take To Tame AI Somewhere along my research of AI, I saw a perfect wording of the basic principle of AI: ‘Garbage in, garbage out’. In other words, if you feed your AI agent poor data, it delivers poor data. It doesn’t filter the given information or complement it by parsing other sources. The principle actually extrapolates onto a business's internal structure (or chaos). Meaning, if your business is chaotic and unstructured, you’re unlikely to finally order it with an AI add-on. The organizational psychologist SynthetIQ Partners CEO Vincenzo Russotto was most articulate on the topic. He explained that companies expect some magic button, some cookie-cutter solution that steps in to organize the chaos. “AI will expand whatever you have,” Vincenzo says. If there is chaos inside, AI will make it worse. Companies need to perform due diligence before they invite AI automation. To some clients, Vincenzo explains, this mandatory audit is a deal breaker, and they vanish—it’s difficult to change the order of things, but reviewing and digging into whys and hows is even worse. Vincenzo Russotto AI will expand whatever you have Going back to what Matt Anderson said about AI and human ability to consider context in decision making, AI can’t be trusted with full automation (yet!) due to not only its lack of consideration for context, but also due to compliance issues and legal and copyright constraints. Human oversight, therefore, is essential for safe and secure AI automation to any extent. Where Is Everything Going? I really liked the comparison Natali Gray gave. She contrasted AI with a calculator: “Everyone feared the calculator when it came into existence, crying about how we'd become stupid or lose our math skills. But, instead, we incorporated it into our math classes, and children ended up learning higher levels of math. I definitely feel that AI (LLMs in specific) is the new calculator”. Everyone feared the calculator when it came into existence, crying about how we'd become stupid or lose our math skills. But, instead, we incorporated it into our math classes, and children ended up learning higher levels of math. I definitely feel that AI (LLMs in specific) is the new calculator Stephen Ellul shares this calculator analogy, predicting a deep next-level corporate AI-powered personalization, ‘adaptive AI’ which learns from company-specific data and behavior, running in the background and making smarter decisions over time. Former VP of Demand Generation & Digital Analytics Malika-Budur Kalila views AI as a powerful time-saving assistant—“irreplaceable” for social media planning, editing, and strategy support—yet believes long-form website content “still needs a strong human element,” since “it’s very easy to identify if it’s human-written or AI-written.” Her argument is that AI is supposed to execute human ideas without replacing human thinking, thus keeping creativity, emotion, and relationship-building at the center of consumer messages. Malika-Budur Kalila irreplaceable still needs a strong human element it’s very easy to identify if it’s human-written or AI-written Conclusion Before I sign off with my own final take on AI, I’d like to summarize the key takeaways of the interviews. The reality is 30/70. Contrary to popular research, in my view, AI only accounts for 30-40% automation.
AI can’t be left unsupervised (yet). A prerequisite for efficient and meaningful productive use is constant refinement and fact-checking. Context, judgment, creativity, and oversight remain exclusively human domains
Incorporating AI ≠ Streamlined operations. AI is just a tool a business uses to free up resources for strategic and creative tasks. It is not the secret ingredient to fixing the internal chaos.
(Loving’ it!) Human talent is replaced not with AI, but with AI-versed human talent. There is no denying that the operational work of copywriters and marketers is easily automatable with AI. If I don’t automate myself, someone else with a proper AI toolkit is likely to take over. This is a very sobering thought—and a good reason to reconsider my professional development. The reality is 30/70. Contrary to popular research, in my view, AI only accounts for 30-40% automation. The reality is 30/70. AI can’t be left unsupervised (yet). A prerequisite for efficient and meaningful productive use is constant refinement and fact-checking. Context, judgment, creativity, and oversight remain exclusively human domains AI can’t be left unsupervised (yet) Incorporating AI ≠ Streamlined operations. AI is just a tool a business uses to free up resources for strategic and creative tasks. It is not the secret ingredient to fixing the internal chaos. Incorporating AI ≠ Streamlined operations. (Loving’ it!) Human talent is replaced not with AI, but with AI-versed human talent. There is no denying that the operational work of copywriters and marketers is easily automatable with AI. If I don’t automate myself, someone else with a proper AI toolkit is likely to take over. This is a very sobering thought—and a good reason to reconsider my professional development. Human talent is replaced not with AI, but with AI-versed human talent. Obviously, AI is going to be a powerful assistant that will accelerate work and ultimately augment life itself. AI per se is neither good nor bad. Its ultimate implication and outcome is in the hands of the user. While AI is worth considering as a full-fledged thinking partner or a critic, it is not a brain replacement. The ominous future where every post, image, and text is like a million others is only possible if we, as humans, collectively abandon our final check position. At the same time, being AI-literate is the new normal. Learning to operate AI is a necessary step for every professional who plans on staying in demand with new clients and businesses. Speaking of which, I’m looking for a comprehensive course on AI. Any suggestions?

The is an opinion piece based on the author’s POV and does not necessarily reflect the views of HackerNoon.

Dana

The AI Reality Gap: What 11 Professionals Revealed About AI at Work That MIT Studies Won't Tell You

About Author

Comments

TOPICS

THIS ARTICLE WAS FEATURED IN

Related Stories

One More Thing in AI: Meta's AI Paradox, Music Copyright Battles, 3D Printing Revolution, and More

Induction Capability: Multi-Token Prediction on Higher-Quality Data

18 AI Marketing Softwares Your B2B Needs to Try Today

3 Ways Artificial Intelligence Will Change Digital Marketing

AI for Marketing Automation: Strategies and Tools for Marketers

AI Search: Exploring the Threat of Traffic Drops and How to Rank on Large Language Models

One More Thing in AI: Meta's AI Paradox, Music Copyright Battles, 3D Printing Revolution, and More

Induction Capability: Multi-Token Prediction on Higher-Quality Data

18 AI Marketing Softwares Your B2B Needs to Try Today

3 Ways Artificial Intelligence Will Change Digital Marketing

AI for Marketing Automation: Strategies and Tools for Marketers

AI Search: Exploring the Threat of Traffic Drops and How to Rank on Large Language Models

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps