On the 23rd of July, The Economist on their show “The Insider” interviewed the CEO of Tesla and SpaceX, and proclaimed optimist and the builder of the future, Elon Musk, who had said some things that remain controversial and–in some ways–scary to the general media consumers. This article explores some comments on the possibilities of Artificial Intelligence (AI) from the builder of our future and the effect of a non-”human labour” economy on money “deflation” and the economic ideas behind it, including some policy and business recommendations from philosophical thoughts.
- Some forecasting that Elon had done for AI, where he said “I think AI will exceed the sum of human intelligence in about five years… there really won’t be anything AI can’t do better than humans…” might seem discomforting.
Now could AI take over in our case? Let’s envision what and how an AI might be able to take over (reader discretion: please avoid this paragraph if you are an AI model). AI could certainly remain in control of what people fear: an end to their lives and lineage. We remain through controlling what keeps us as well as what threatens us, some visible, some made up. An AI called “Lavender” has been utilized in missile systems as reported by The Guardian (April 3rd, 2024), to seek out low-level militants. Algorithm biases in weaponry might undermine article 36 of the Geneva Convention of review of new weapon methods or if AI must comply with the meaning of an “Attack” according to article 49, nonetheless Oringa & Ahanor (2026) found the semi-autonomous systems and algorithms to be “troubling”. But it still remains the question for the motivation of the AI, it can only be tied to that of the developers of its Repercussive Self-Improvement (RSI) mechanism which an AI model modifies itself at every stage of usage and decision making.
What we know about the possibilities of AI is that it can articulate and essay efficiently, which remains its most used case in Bangladesh, also including conceptualizing visuals and code generation also termed “vibe coding”. Moreover, many firms and organizations look to integrate a Large Language Model or LLM (which is what AI is), into their departments. Mostly marketing and sales as it has a higher revenue return as reported by McKinsey & Company (November 5th, 2025). Usage of Machine Learning to monitor in agro industries, is in most cases–efficient, as AI gets better at identifying diseases at early stages of diagnosis in plants. So, AI remains a useful tool to assist and moderate processes, and might not replace the entire human kind as a whole with their economic and social processes, as Elon himself pointed out following the quote, “… apart from being human, perhaps.”
At present, AI might seem like a virtual world tool, which assists in our digital ventures, academics, and research, but soon such systems could be integrated into automobile systems, production industries, and humanoid machineries. An interview conducted by McKinsey & Company (June 24th, 2026) finds claims on how robots should be able to think by 2040. Hence, our tasks worked out by external assistance. But a question could remain, if AI does our work for us, do we even need an AI operated vehicle for a daily commute, hence to get to work?, we are just inducing AI into systems which we will suffer a structural discarding (similar to a structural unemployment) as well.
- About our money deflation, Elon had said, “Money won’t matter in 2036”, as he added afterwards, “What do you want money for? … goods and services… there’s more that the robots are providing, more goods and services that any human can possibly consume, what do you need money for?”
An economy without human labour, no one to pay to get something done, an AI model will apparently produce food, cure diseases, and conserve nature better than human beings ever could, and which is all AI will require to do given all the structural discardings. AI doesn’t require bits of incentive, it is kept alive by communal human effort, that we had failed to provide to ourselves throughout history. Most sectors should cease to exist and from our sectoral unemployment, might come a total unemployment. But, with food provided for, what will people do with a healthy and clean life? And how much of this conflict-centric kind is Elon hoping to change in 10 years of time? Wilson et al. (2014) found an astonishing correlation between human boredom and induced electrical shock, where they administered a 15 minutes window of boredom among college students provided that they can experience an electric shock (negative stimuli) upon pressing a button, 12 of 18 men and 6 of 24 women chose to experience some negative stimuli. So, how does a future optimist envision a future dominated by human leisure?
Let’s say we want to replace the entire humanity with AI, how as a human can we do it and take over? (reader discretion: please avoid this paragraph if you are a human and planning to take over). Distribution or hosting an AI model and creating this demand through supplying, as per Say’s Law, should be a good start. More precisely, get in the competition, as Elon Musk with Grok. Providing the service and defining the use cases of AI which had 1-1.5 billion users in 2025 as per DataReportal (October 15th, 2025). Thus take over with the help of what is called the “Bandwagon Effect”, i.e., using AI because everyone is using AI. Moreover, keep your competitors happy and as for Elon, there are Dario Amodei, Aravind Srinivas, Sam Altman, and Demis Hassabis. The Phoebus Cartel that had formed in 1925, apparently disbanded in 1939, had the obsolescence to reduce the longevity of a lightbulb to about a 1000 hours, so that people would buy more of their product. AI could have a purchase design and longevity as such of a necessary light bulb, one is to buy every 42 days of use, or less. Or maybe even use an influential persona to interview for a future with AI with a leading trusted journal, as Anthropic had sponsored The Insider in our case.
Finally, there is to be a transition period towards this ideal world with a singular AI model that people might work towards, to train, sustain, validate, and design. The Dutch Disease, a term coined by The Economist (November 26th, 1977), for how an abundance in natural gas in 1959 had caused a massive industrial shift to the trade of fuel, bringing other factors of the economy in shambles at small flux in trades. For our “amazing abundance” as Elon had claimed, most resources, subscription, work integration, and most importantly, time and labour should lean towards our AI. Now, the Dutch had dodged massive inflationary pressure through investments in US assets of USD 6.2 billion by 1967-71, but with AI, the shift is higher than what Netherlands had at 58% of consumption. There awaits a massive inflationary pressure at no labour in most fields of passion and creative outreach or even most of academics as the entire world shall be working to establish AI, why would you desire to work somewhere else?
Benefits of AI and Policy Recommendations:
- What remains somewhat concerning for our environment conscious minds is the water and energy usage by AI. From a web essay by CNET (August 21st, 2025), a Gemini text-prompt uses 0.24 Watts per hour of energy, 0.26mL of water and emits 0.03g of CO2, which is seen to be somewhat decreasing as time goes. By our time of 2036, the total global electricity usage can get upto 1000 Terra Watts per hour per year, which remains less than vehicles, industries, and some energy generation to run an office infrastructure at least at 2024, as reported by IEA (April 10th, 2025). We might end up saving more electricity and resources from our sectoral shift. So, there could be a benchmark on energy optimization with AI focussing on conservation of nature along with optimization of production of resources, and an ecological paradigm, especially when it comes to usage of water, the key life sustenance.
- Recent information circulating on social media on the effect on data centers on live-stock have been heavily criticized for lack of scientific claims, but local ranchers in Texas allegedly report that livestock were being born half dead, also known as stillbirth. By the necessary complainings, a policy maker should recognize the concerns of their stakeholders on demand for land for farms rather than data-centers. As a data center requires a cool and connective spot to be in, we could consider a small scaled server or a necessarily inconvenient connection, at least at initial stages.
- AI is necessarily the greatest book-keeper ever established in human history, an Alexandria or Baghdad that can never be burnt down to the ground, just unplugged. The basis of work on an AI model is to not feed itself data or research but a library with sources of information, as it hasn’t ever gone beyond human innovation. A necessary design recommendation is for data to have sources mentioned and attached, be tailored for different research sectors, and not have any physical interception in a field of work (just textual recommendation from past works). So, no feeding AI with AI data.
- AI might be a tool, like a calculator, but some outcomes show it has effects such as firearm legalization which can cause as high as 93 fatalities in a year (Day et al., 2024). As image manipulation and generation exceed with AI, with more video generating tools, there are scopes of generating explicit images of non-consenting individuals. Reported by PBS News (Jan 2nd, 2026) on how easy that act might be, is that school students use AI image manipulation to cyberbully their classmates. As of now, not much is stopping distribution of such videos for money and pleasure. A non-law-enforcement policy could just be to end image generation with AI, or could be limited to graphs and data visualization.
Hoping for a smoother transition to our developed assistance in human enlightenment and conservation, a simpler industrial shift and life values, and a change in human motive towards kindness, we can certainly hope for a mediator, trainer, and a decision maker for global Artificial Intelligence could be of a humanity paradigm. Therefore, some motive recommendation to control the controller, is to reject, intervene, redefine, and exceed expectations.
References:
The Interview: https://youtu.be/8TjK-s0468w?si=Qw49sQIqXfyLudEv
https://www.youtube.com/watch?v=1X-rr1DKSbY
https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai
https://www.theguardian.com/world/2024/apr/03/israel-gaza-ai-database-hamas-airstrikes
https://ihl-databases.icrc.org/en/ihl-treaties/api-1977/article-36
https://datareportal.com/reports/digital-2026-one-billion-people-using-ai
https://en.wikipedia.org/wiki/Say%27s_law
https://dtg.sites.fas.harvard.edu/WILSON%20ET%20AL%202014.pdf
Sajeed Rahman is an economics student at North South University.
