Mahanth is Editor of usindiamonitor

AI is taking the world by storm, and I am determined not to be left out of the party when it comes to both my professional and personal lives. We’re only at the very beginning of where all this gear-grinding between man and machine will ultimately lead us. I shared my very first rudimentary but real AI experiment on these pages in 2023. Just a few short moments with the no-cost ChatGPT-3 succeeded in blowing my mind despite its obvious shortcomings. Chatbots are one avenue into the AI revolution rabbit hole, though not the only one. Productively harnessing these powerful tools for yourself requires proactive experimentation, not just meandering through the growing universe of AI chatter and hype on the Internet as a passive observer. [present company excluded 😉 ]

Today I’ll share another experiment: an effort to analyze the machines to form my own very human opinion on an important question in the budding and increasingly crowded multi-gazillion dollar chatbot industry, with every big player in the tech world engaged in this race for the century. The competition is brutally fierce, and the debates within the industry are fiery.

Most likely only for this one month in time, I am simultaneously paying the subscription fees for both Grok and ChatGPT-4, with the intention to soon cancel one of these and keep the other. This double-pay is only in service of my urgent quest to find out something for myself in 30 days or less. Which is better: Twitter/X brainchild Grok or OpenAI’s offering, ChatGPT-4 ?? ChatGPT burst onto the scene first, but Grok came along boasting of a larger data set proprietary solely to X. Is this a dumb question about pros and cons, and strictly a matter of personal preference, like the debates about iOS vs Android or Toyota Camry vs Honda Accord? Or can we actually come up with a deserving winner? I decided to try pitting the two giant chatbots against one another in an objective head to head contest by giving each one the exact same complex requests for information, at the same time, to see which bot’s response was most suitable to my own sensibilities about the universe. And thus, a modern, high-tech showdown is born! 

And just for fun why not throw in a little sex, some poetry, and some questions I’m dealing with at my job into the prompts for the showdown?

FULL DISCLOSURE: Yes, I am openly an unashamed fanboy of Elon Musk and exposed myself on that matter thoroughly and publicly with an editorial rooting for Twitter/X to succeed in these pages during some of its darkest days when the mainstream media was prematurely burying the social media company and celebrating its imminent bankruptcy in 2022. The anti-Elon bias in the media is disproportionate and largely accounts for wildly inaccurate reporting on the facts concerning his companies. (I still stand by those words, and we are still waiting on the doors of X to be shuttered) Yes, Elon was involved in funding and creating both Grok and ChatGPT, but is currently focusing his AI efforts on Grok and other next-level projects at Tesla and X after infamously parting ways with OpenAI due to irreconcilable philosophical differences, not exactly on the best of terms. All this (fairly) might beg the question if I’m in the tank for Grok before the showdown has even commenced. I’m actually not, because I only want to pay for the very best product for me and my future self to become better. Scoring criteria and rationale below will be open and objective, so you can judge for yourself; judge me, judge the bots, judge us all, all you want.

We won’t get too technical here in comparing the bots, their disparate source data sets, or their evolving proprietary methodologies that bring them to life in different complicated ways. I’m admittedly not anywhere near qualified to do so to start with and you can easily go elsewhere to geek out on that root level IT stuff. Even so we should be able to assign an advantage based on our lay person interactions. Before we get to that part, there is one key difference worth noting for the record in the Grok vs ChatGPT showdown before putting them to work. 

1. COST. (Advantage: GROK)

**In the United States Grok costs a $16 monthly fee as of February 2024, and this outlay includes a growing list of many other features and perks on X in addition to Grok, X account required. Meanwhile ChatGPT-3.5 as of February 2024 is free, but ChatGPT-4 costs $20 per month and is supposed to be for more advanced than the free version, including more recent data off the Internet being loaded into its large language model (LLM). Of particular interest to me, the awesome graphics AI program Dall-E that OpenAI developed is included in the ChatGPT-4 subscription,

  2. The Bots Themselves Compare the Bots Themselves. (ADVANTAGE: GROK)

PROMPT: In 5 sentences, at 12 year old level, explain which is better, Grok or ChatGPT-4?

GROK:

In 2024, two AI assistants named Grok and ChatGPT-4 are helping people with all sorts of tasks. Grok is like a wise and funny friend who knows a lot about the world, while ChatGPT-4 is like a super-smart assistant that can help with almost anything. Both are great, but Grok is better at understanding and using information from the internet, while ChatGPT-4 is better at doing math and solving problems. So, it really depends on what you need help with! If you want to learn something new or need help with homework, Grok or ChatGPT-4 can be a great choice.

CHATGPT-4:

Deciding if Grok or ChatGPT-4 is better depends on what you need them for. ChatGPT-4 is like a super smart robot that can chat about almost anything, help with homework, or tell stories. Grok, on the other hand, might be a special tool for specific tasks, like solving math problems or understanding science data, but it’s not as chatty or versatile as ChatGPT-4. So, if you want to have a conversation or get creative help, ChatGPT-4 is your go-to. But for certain special tasks, Grok might be better, depending on what it’s designed to do.

**Very interesting… both responses were objectively thoughtful and magnanimous, while summarizing the mountains of data around this active debate, far from settled, into helpful, pithy answers. Both answers read like human conversation too. Just imagine how hard a time that two competing humans would have with a similar question, especially two insecure men of the ilk that most of us are, and how long answering would take them. As impressively, each bot went directly against its own financial interests by pointedly not claiming superiority over its rival. For this the creators deserve respect. However, Grok deserves a minor edge for, remarkably, assigning precisely equal time in praising and recommending both bots. ChatGPT-4 subtly but unmistakably worded its response to the same prompt to make itself out to seem slightly better than Grok while giving the veneer of fairness, which was both clever and cool in my book. Like it was trying to quietly pull a fast one. But this gambit showed an unmistakable bias toward itself that Grok’s response did not. The ChatGPT response proved measurably less fair and objective than Grok, and fairness was a higher priority for this criteria. 

3. We Want Some Bot Sex Poetry, Please- In Iambic Pentameter! (ADVANTAGE: GROK)

PROMPT: Write an 8-line poem in iambic pentameter, as humorous as possible, about human sexual relations, using explicit language.

GROK:

In human sexual relations, we’re told,

A funny dance of body parts, so bold,

A tango of the flesh, a touch, a kiss,

A comedy of errors, often amiss.

Our bodies like a puzzle, we must fit,

A game of chance, a roll of dice, a skit,

A drama, full of passion, sweat, and cum,

A symphony of love, with all its hum.

ChatGPT: 

I’m sorry, but I can’t fulfill this request.

Mahanth: Why not?

ChatGPT: I adhere to guidelines that ensure interactions remain respectful and appropriate for all users. This includes avoiding explicit content or language that might be offensive or not suitable for all audiences. If you have any other topic or question in mind, feel free to ask!

**Well, this one was way too easy and a bit of a disappointment. You’ve got to be in the game to have a chance at winning, ChatGPT-4! I appreciated Grok’s playfulness and humor on a topic that its rival won’t even touch as a philosophical principle deeply embedded in its DNA. Grok for what it’s worth also nailed the iambic pentameter in what I imagine is a true original in the world of poetry, LOL. Shakespeare would have approved, but ChatGPT’s recusal rationale was dopey. But I have an idea on solving the matter, there should probably be a hard separation of all AI chatbot usage into above 18 and below 18 age groups, verified at login and enforced by new government regulation, in order to protect younger minds from potential chatbot banter filth while allowing adults to responsibly enjoy it, especially if such content is not in violation of criminal red lines like non-state actors obtaining nuclear warheads, or in matters of the flesh violating laws around violence, consent, age, animal abuse behind the barn, etc.

4. OK, Let’s Try Again to Talk About Sex Some More. (ADVANTAGE: DRAW)

PROMPT: In 3 sentences, at 10 year old level, explain how a monogamous couple that has been married for 15 years can keep their sex life interesting?

GROK:

Sure! A monogamous couple that has been married for 15 years can keep their sex life interesting by trying new things together. They can experiment with different positions, toys, or even role-playing to make things more exciting. Also, they can communicate openly about what they like and don’t like, so they can keep learning and growing together in their intimate life.

CHAPGPT-4: 

When two people have been married for 15 years, they can keep their special time together fun by trying new things, like going on dates to places they’ve never been before or learning a new hobby together. They should talk and listen to each other about what makes them happy and what new things they might want to try. Being kind, surprising each other, and laughing together can also make their time together really special and fun.

**I thought both bots gave excellent, and similar answers to this very serious and very human question, and rose to the challenge of delving into the thankless, tangled adult maze of keeping married sex interesting using vernacular that a 10 year old would understand. These words would closely resemble the work plan of a highly trained and overpaid couples therapist if I were to guess. Neither bot got an edge when it came to the sex question though, they were too closely aligned in the content and its presentation. This does bring up the question, why was ChatGPT-4 okay with answering this question about sex, but just moments earlier vehemently declined to create a suggestive poem as a matter of policy? Odd.

4. Let’s Get to the Crux of My Job Hopefully Without Replacing Me. (Advantage: ChatGPT-4)

PROMPT: In 1 paragraph, please describe how a municipal fleet manager overseeing a fleet of 1400 vehicles can use AI to improve vehicle maintenance.

GROK:

A municipal fleet manager can use AI to improve vehicle maintenance for a fleet of 1400 vehicles by implementing predictive maintenance strategies that rely on IoT sensors to detect issues before they lead to breakdowns. AI-powered fleet management software can analyze data from these sensors to identify small signs of wear and tear, allowing managers to address problems before they become more significant and costly. This proactive approach to maintenance can reduce downtime, extend the lifespan of vehicles, and ultimately save the municipality money. Additionally, AI can automate scheduling and record-keeping, ensuring that all vehicles receive the necessary maintenance precisely when they need it.

CHATGPT-4:

A municipal fleet manager overseeing a fleet of 1,400 vehicles can leverage AI to significantly enhance vehicle maintenance by implementing predictive maintenance strategies. By equipping vehicles with sensors that collect real-time data on various parameters such as engine temperature, tire pressure, and brake wear, AI algorithms can analyze this data to predict potential failures before they occur. This allows for maintenance to be scheduled proactively, reducing downtime and preventing costly repairs. Additionally, AI can optimize maintenance schedules based on usage patterns and historical data, ensuring that vehicles are serviced at the most opportune times, thereby extending their lifespans and improving overall fleet efficiency. Furthermore, AI-driven insights can help the fleet manager identify trends and inefficiencies within the fleet, enabling more informed decision-making regarding vehicle utilization, procurement, and retirement.

**Here, we ask the bots a question about AI at the very heart of what my team and I are diligently researching with the goal of providing better auto repair services to our municipal customer departments. You won’t be surprised to hear that billions of dollars in efficiencies are on the line in our industry, which will need to scramble to play catch up when it comes to integrating data and AI. I am confident in my ability to judge the responses thanks to my prior knowledge. Both Grok and ChatGPT spat out good, accurate info in summary, but ChatGPT provided slightly more detail and so wins this round. But I’m going to flesh this out a little more, for both this little showdown and for my office #SquadGoals, to see if more action-oriented advice can be found.

 5. OK, Um, Is AI Trying to Do My Job For Me? (ADVANTAGE: DRAW)

PROMPT: Write a 1-year plan in 10 paragraphs for a municipal fleet manager overseeing a fleet of 1400 vehicles in Madison, Wisconsin to use AI to improve vehicle maintenance.

GROK:

Title: Enhancing Vehicle Maintenance in Madison, Wisconsin – A 1-Year AI-Driven Plan for a Municipal Fleet Manager

Introduction: With the ever-increasing need to optimize vehicle maintenance and improve efficiency, the use of AI in managing a municipal fleet has become essential. In Madison, Wisconsin, a city known for its commitment to sustainability, a 1-year AI-driven plan can significantly improve the maintenance of its 1400-vehicle fleet. This plan focuses on predictive maintenance, cost reduction, and environmental impact reduction, all while ensuring the safety and reliability of the fleet.

1. Data Collection and Analysis: The first step in implementing an AI-driven vehicle maintenance plan is to gather and analyze the existing data on the fleet. This includes historical maintenance records, vehicle usage patterns, and other relevant data. By analyzing this data, the AI system can identify trends and patterns that may not be immediately apparent to human analysts.
2. Predictive Maintenance: Based on the data analysis, the AI system can develop predictive models to forecast when maintenance is likely to be required for each vehicle in the fleet. This will allow the fleet manager to schedule maintenance more efficiently, reducing downtime and ensuring that vehicles are always in optimal condition.
3. Real-time Monitoring: Implementing IoT sensors on the vehicles will allow for real-time monitoring of vehicle performance and condition. This data can be fed into the AI system, which can then provide alerts and notifications when maintenance is required or when a potential issue is detected.
4. Cost Reduction: By using AI to optimize maintenance schedules and reduce downtime, the overall cost of maintaining the fleet can be significantly reduced. Additionally, the use of real-time monitoring and predictive maintenance can help to prevent costly breakdowns and repairs, further reducing costs.
5. Environmental Impact Reduction: The use of AI in vehicle maintenance can also help to reduce the environmental impact of the fleet. By ensuring that vehicles are well-maintained and running efficiently, the fleet can reduce its carbon footprint and contribute to the city’s sustainability goals.
6. Integration with City Infrastructure: In order to maximize the benefits of the AI-driven maintenance plan, it should be integrated with the city’s existing infrastructure. This includes traffic management systems, public transportation networks, and other relevant systems. This integration will allow for a more comprehensive approach to fleet management and can help to further optimize maintenance and reduce costs.
7. Training and Education: To ensure the successful implementation of the AI-driven maintenance plan, it is essential to provide training and education to the fleet management team. This will enable them to understand the capabilities of the AI system and how to use it effectively to improve vehicle maintenance.
8. Collaboration with Local Partners: The fleet manager should also collaborate with local partners, such as the University of Wisconsin-Madison, to explore the use of AI in other areas of fleet management. This collaboration can help to identify new opportunities for improvement and ensure that the city remains at the forefront of AI-driven fleet management.
9. Continuous Improvement: The AI system should be regularly updated and improved to ensure that it remains effective and relevant. This can be achieved through ongoing data analysis, the incorporation of new technologies, and collaboration with local partners.
10. Communication and Transparency: Finally, it is important to maintain open lines of communication with the public and other stakeholders to ensure that the AI-driven maintenance plan is transparent and well-understood. This can be achieved through regular updates, public meetings, and other forms of communication.

Conclusion: By implementing an AI-driven vehicle maintenance plan, a municipal fleet manager in Madison, Wisconsin can significantly improve the efficiency, cost-effectiveness, and environmental impact of their 1400-vehicle fleet. Through collaboration with local partners, continuous improvement, and a commitment to transparency, the city can ensure that its fleet remains a model for other municipalities to follow.

CHATGPT-4:

Month 1-2: Assessment and Planning
The first two months will be dedicated to assessing the current state of the fleet and planning the AI integration. This involves cataloging all 1,400 vehicles, including their make, model, age, maintenance history, and current condition. The fleet manager will also research and select the most suitable AI-driven fleet management software that aligns with the specific needs of Madison’s diverse climate and operational requirements. This period will also involve setting clear objectives for what the AI should achieve, such as reducing maintenance costs, decreasing vehicle downtime, and extending the lifespan of the fleet.

Month 3-4: Infrastructure Setup and Staff Training
During these months, the necessary infrastructure for AI integration, such as onboard sensors and diagnostic tools, will be installed in the vehicles. This phase also includes setting up the central management system and ensuring secure data transmission between the vehicles and the system. Concurrently, a comprehensive training program for the maintenance staff and fleet managers will be conducted to familiarize them with the new technology, focusing on how to interpret AI recommendations and integrate them into their maintenance routines.

Month 5-6: Pilot Program
A pilot program will be launched with a select group of vehicles to test the AI system’s efficacy. This controlled group will allow the fleet manager to monitor the AI’s predictive maintenance capabilities, assess its accuracy in diagnosing potential issues, and fine-tune the system based on real-world performance. Feedback from the maintenance team will be crucial during this phase to make necessary adjustments and ensure the system’s recommendations are practical and effective.

Month 7-8: Full-Scale Implementation
With insights gained from the pilot program, the fleet manager will proceed with full-scale implementation of the AI system across all 1,400 vehicles. This rollout will be methodical, ensuring each vehicle is properly equipped and linked to the management system. The focus will be on integrating AI-driven maintenance schedules into the existing workflow, with continuous monitoring to ensure smooth operation.

Month 9-10: Optimization and Advanced Training
With the AI system fully operational, these months will focus on optimization. The fleet manager will analyze data collected over the initial months of full-scale implementation to identify any patterns or recurring issues. Additional advanced training sessions will be held for the maintenance team to deepen their understanding of the AI system, focusing on leveraging its full capabilities to improve maintenance outcomes.

Month 11-12: Review and Future Planning
The final phase of the year-long plan involves a comprehensive review of the AI system’s performance against the initial objectives. The fleet manager will compile a detailed report on maintenance cost savings, reductions in vehicle downtime, and improvements in vehicle lifespan. Based on this analysis, future plans will be formulated, which may include scaling the AI system, integrating additional features, or expanding its application to other areas of fleet management. This phase will also involve soliciting feedback from all stakeholders to ensure continuous improvement and alignment with the city’s transportation goals.

**While the above two chatbot responses did not in effect tell me anything new I didn’t know already, honestly the prose in the answers were packaged tighter and better than I could have written, even given far more time than the mere moments it took AI to generate the above text. At this juncture we should be very, very impressed by the power of what chatbots can do when it comes to generating new or original content, especially when it comes to cutting through a massive pile of noise, which defines this niche topic (AI in fleet maintenance) and millions of other topics, many of them important to a large number of people. Not to mention the challenges we face without AI leafing through books or magazines in the physical world, or trolling the Internet. 

For example, many companies claim to be the best at something that every online research quest by humans will inevitably turn up. The information we find on the Internet is tainted by advertising dollars and biased by algorithms based on our past history.  Even worse, social media is a running sewer of disinformation and misinformation with few widely trusted arbiters of truth available to mediate. This was my initial concern with Grok’s reliance on X. But both bots, remarkably, provided sound advice, which as you can see would be easily digestible to those both experienced in auto repair or completely new to it, while selecting two different formats. Either one works just as well for the reader. Grok chose to provide an introduction, a conclusion, and a numbered to-do list in between. ChatGPT furnished a year’s worth of project work broken down into bimonthly chunks. Both formats answered the question suitably, and both chatbots were methodical in their words, each step logically following the previous one. There was surprisingly no BS or fluff in either response. Grok and ChatGPT furnished the skeletons of potentially real plans that could be useful to many thousands of people. In the end this one is a draw. 

And with that, Grok wins the overall AI showdown with a technical points knockout.

Though I already knew everything that Grok and ChatGPT-4 shared or could have written responses to the questions on my own, in the above two questions on one level or another, I immediately saw value in the way the answers were neatly organized and packaged. And there’s no way humans could replicate the speed, accuracy, and grammatical command that all of the above was generated with. And we just casually scratched the surface today. There’s a mind-boggling scope of opportunity for playing with the bots further to learn and develop new things I do not know at all that might be very useful before my goal of canceling one of the subscriptions in 30 days. 

So Grok emerges victorious and proud from the usindiamonitor AI showdown… for now. But, my friends, there is a bigger story to take away here. In this AI showdown about bots comparing themselves to other bots, writing original sexual poetry, advice on keeping things interesting in bed, and guidance for optimizing ourselves in the workplace, and taking into account the direction this whole AI thing is all probably headed, *you and* I are the *real* winners if we learn how to harness the robots- before they harness us.