ChatGPT experiment: oil price forecasts for 7, 14, 30, 60 and 90 days
A ChatGPT experiment predicted oil at 90 days in March 2026: from $106 down to $90. The thread was lost in the noise.
English · Original discussion in Spanish · Published
ChatGPT predicts oil: a 90-day forecasting experiment that faded into the noise
On 20 March 2026, an artificial intelligence experiment produced a series of oil price forecasts for 7, 14, 30, 60 and 90 days. The ranges: $104-109 for 27 March, gradually falling to $88-93 on 18 June. The central prices: $106, $103, $99, $96 and $90 respectively. The implied trajectory: a tight market in the short term, partial normalisation in May and cooling towards June. What promised to be a test of AI's predictive ability turned into a sterile discussion with no trinc-up and no verification.
The AI's bet: a downward trend with nuances
The ChatGPT model projected a gradual price compression from $106 in one week to $90 in three months. A quick reading suggested the market was still tight at the end of March, but that the fundamentals would allow a pogre easing. The narrow ranges suggested moderate confidence in the direction, though not in the exact levels. However, the exercise was left unfinished: no user provided any actual later data to confirm or refute the predictions.
What the noise swept away
The thread degenerated into personal attacks and memes, obscuring any substantive debate about AI's reliability in volatile markets. There was no analysis of the underlying factors: OPEC+ supply, Chinese demand, strategic reserves or seasonality. The experiment sank before it could be evaluated, leaving an open question: is ChatGPT any good at forecasting commodities, or is it just a stylistic exercise?
Related discussions in the forum
On 20 March 2026, an artificial intelligence experiment produced a series of oil price forecasts for 7, 14, 30, 60 and 90 days. The ranges: $104-109 for 27 March, gradually falling to $88-93 on 18 June. The central prices: $106, $103, $99, $96 and $90 respectively. The implied trajectory: a tight market in the short term, partial normalisation in May and cooling towards June. What promised to be a test of AI's predictive ability turned into a sterile discussion with no trinc-up and no verification.
The AI's bet: a downward trend with nuances
The ChatGPT model projected a gradual price compression from $106 in one week to $90 in three months. A quick reading suggested the market was still tight at the end of March, but that the fundamentals would allow a pogre easing. The narrow ranges suggested moderate confidence in the direction, though not in the exact levels. However, the exercise was left unfinished: no user provided any actual later data to confirm or refute the predictions.
What the noise swept away
The thread degenerated into personal attacks and memes, obscuring any substantive debate about AI's reliability in volatile markets. There was no analysis of the underlying factors: OPEC+ supply, Chinese demand, strategic reserves or seasonality. The experiment sank before it could be evaluated, leaving an open question: is ChatGPT any good at forecasting commodities, or is it just a stylistic exercise?
Related discussions in the forum
- 5 tonnes of HORO sent to Dubai by Plus-ultra blew the case open · 119 replies
- Gorpe de hestao eh un gorpe de hestao · 39 replies
- Latin discovers, horrified, another repulsive Spanish scourge: THE LITTLE SPANIARDS EAT WAY TOO MUCH BREAD AND VERY LITTLE RICE, LIFE ONLY TASTES GOOD WITH THE ARABS · 39 replies
- The head of the ETA terrorist group dies in Venezuela, where Maduro had made him an adviser to PDVSA · 32 replies
Also available in: Deutsch
Summary of a discussion on Burbuja.info - Foro de economía, actualidad y política., translated from Spanish and reviewed before publication. Read the full discussion (6 replies).
More summaries
-
Roeben Gisbert: no oil exists and prices aren't falling
The claim that oil does not exist reopens the peak oil debate: what is running out is not crude itself, but the technical capacity to extract it.
-
Saudi Oil Production Hits 1990 Lows as Red Sea Rerouting Drives Prices to $107
Crude oil trades near $107 per barrel as Saudi Arabia ramps up production to 1990 levels and shipping routes shift around the Cape of Good Hope.
-
Stocks and oil: the strange dance nobody can explain
Announcements of a flood of crude pointed to a crash, yet tech stock indices are hitting all-time highs. Where are investors seeking shelter?
-
Electric Car Fire in Garage Forces Building Reinforcement
A rental electric vehicle caught fire in a garage, destroying an adjacent car. Residents stayed in shelters while the building required emergency structural reinforcement.
-
Viral tweet falsely puts 'invasion' migration quote in Trump's mouth
A viral tweet attributes to Trump a call on migration in Europe an 'invasion' and labels the United Kingdom 'suspect number one'. The post offers no evidence.
-
Peanut pay, Swiss-quality output: how Spain's tech giants profit from cheap engineering
Indra reported a net profit of 436 million euros in 2025. Its senior engineers report gross salaries of 32,000 to 41,000 euros a year on Glassdoor.