top of page
  • Writer's pictureWilliam Webster

Is Owning an EV Really Cheaper? GPT-4 Analyses the Costs Compared to Your Old Car

Updated: Nov 4, 2023


Can GPT-4 Assist in Making Electric Vehicle Purchase Decisions?

Purchasing a new car is undeniably a significant financial commitment, often ranking as one of the heftiest expenditures in a household's annual budget. In recent times, there's been a growing shift towards environmentally friendly transportation alternatives, with electric vehicles (EVs) leading the charge. This begs the question: Is it financially sound to make the switch to an EV from an older, conventional vehicle? Here, we'll delve into a personal exploration of this question using the capabilities of GPT-4, OpenAI's advanced language model.


Understanding the Model's Calculations: For those eager to know the answer jump to the conclusion below. However, understanding the journey to that answer provides a deeper comprehension of the intricacies of such models.

When initially tasked with calculating the cost comparison between a traditional car and an EV, GPT-4, despite its prowess, made notable miscalculations. For instance, it erroneously combined the initial purchase cost of the vehicle with its depreciation value. In another instance, it misjudged the petrol cost calculation. Such errors underline the importance of constructing clear and connected prompts for the model.

After refining the prompt, here's a summarised look at the findings:

  1. Without Tax Incentives: Keeping the older vehicle proved more cost-effective when solely considering basic expenses.

  2. With Tax Incentives (Method 1): After incorporating available tax reliefs for EVs, the financial difference between retaining the old car and purchasing a new EV became marginal.

  3. With Tax Incentives (Method 2): A different approach to applying tax incentives tipped the scales in favour of the EV, highlighting the substantial impact of government policies on such decisions.

Here's the discussion:










Here are the key takeaways from the discussion:

  • Prompt Design: Crafting concise, interconnected prompts improves the accuracy of GPT-4's outputs. For instance, mentioning fuel consumption followed immediately by its cost can guide the model better.

  • Model Reliability: While GPT-4 can occasionally err, it undeniably offers a valuable tool for dissecting complex financial scenarios. However, it's prudent to possess some foundational knowledge about the subject matter to discern the model's accuracy.

  • Accuracy versus approximation: GPT- 4 is convincingly accurate but it's better to take its answers here as approximations that shed light on what is a complex interplay of issues. Nevertheless, it acts as a good guide.

  • As an Aside - Government Policies: Government incentives, especially tax reliefs, can play a pivotal role in making EVs a more viable financial option for individuals who can claim tax relief, for those who can't the numbers aren't so good.

Conclusion: The transition from relying on conventional wisdom to harnessing advanced AI models for complex decision-making is already underway. As more individuals realise the potential of tools like GPT-4, questions once deemed too intricate to tackle might soon become commonplace inquiries.


Still, it's essential to approach such models with a critical mind and to ensure that the prompts provided are clear and connected.

With the right approach, GPT-4 can indeed offer insightful perspectives, making it easier for consumers to make informed choices about their next vehicle purchase. This is a democratisation of information putting power in the hands of those who can use it.


8 views0 comments

Recent Posts

See All

Comments


bottom of page