How many conflict deaths will there be in Sudan in 2025? In this question, Jeremy’s Multi-AI Oracle is a side competitor with the machine forecasting bots of the VIEWS competition.
Model value:
* Less than 1000: 5
* Between 1000 and 3000: 15%
* Between 3000 and 5000: 35%
* Between 5000 and 8000: 25%
* Between 8000 and 12000: 15%
* More than 12000: 5%
Obtaining outside newsfeed. News from various sources: Based on the most recent news articles, here’s a summary of the current situation in Sudan:
Most urgently, multiple sources are reporting a devastating airstrike by the Sudanese army on the Tur’rah market in North Darfur, located about 80km north of el-Fasher. The attack, which occurred on Monday night (March 25, 2024), has reportedly killed hundreds of civilians. This incident represents one of the deadliest single attacks in the ongoing conflict.
The broader context shows an intensification of fighting in Sudan’s capital Khartoum, where the Sudanese Armed Forces (SAF) recently claimed to have retaken the Republican Palace from the Rapid Support Forces (RSF). The conflict, which began in April 2023, has resulted in over 28,000 deaths and created a severe humanitarian crisis. Current reports indicate that civilians continue to be targeted in both artillery shelling and aerial bombardments.
The UN and other international organizations are warning of a deteriorating situation, with concerns about potential renewed civil war. The conflict has led to widespread displacement, with many civilians fleeing to neighboring countries. The situation is particularly dire in Khartoum, where ongoing battles have left many neighborhoods in ruins and created severe challenges for humanitarian access and civilian safety.
Summarizing outside newsfeed:
“Recent reports indicate a concerning escalation in the conflict in Sudan, particularly following a devastating airstrike on the Tur’rah market in North Darfur, allegedly carried out by the Sudanese army, resulting in significant civilian casualties. This incident is considered one of the deadliest since the conflict began in April 2023. Fighting in Khartoum has intensified, with the Sudanese Armed Forces claiming control of the Republican Palace from the Rapid Support Forces. The conflict has led to a severe humanitarian crisis, over 28,000 deaths, and substantial displacement, as civilians face targeted attacks and flee to neighboring countries. International organizations are warning of a potential descent into further civil war, noting especially difficult conditions in Khartoum that hinder humanitarian aid and compromise civilian safety. Caution is advised given the potential presence of unverified information in these reports.”
Fact-checking news summary: Here is a list of specific facts stated in the summary, along with evaluations of their correctness, importance, and relevance:
1. **Fact:** Recent reports indicate a concerning escalation in the conflict in Sudan.
– **Correctness:** True (based on general awareness of the situation in Sudan, there has been an escalation in conflict).
– **Importance:** Important (sets context for understanding future conflict developments).
– **Relevance:** True (relevant, as escalation impacts future conflict deaths).
2. **Fact:** An airstrike on the Tur’rah market in North Darfur was allegedly carried out by the Sudanese army.
– **Correctness:** Needs verification (specific details may require verification from authoritative sources since the information is described as unverified).
– **Importance:** Important (illustrates recent conflict intensity and possible perpetrators).
– **Relevance:** True (relevant, as army actions impact conflict trajectory and casualties).
3. **Fact:** The airstrike resulted in significant civilian casualties.
– **Correctness:** True (consistent with reported severe impacts of such airstrikes).
– **Importance:** Important (critical to assessing victim counts and future conflict predictions).
– **Relevance:** True (relevant, as it influences projections of future deaths).
4. **Fact:** This incident is considered one of the deadliest since the conflict began in April 2023.
– **Correctness:** Likely true (without specific numbers, “one of the deadliest” might be subjective but is plausible given the described disaster).
– **Importance:** Important (contextualizes the severity of the conflict).
– **Relevance:** True (relevant to predicting future conflict deaths based on past incidents).
5. **Fact:** Fighting in Khartoum has intensified.
– **Correctness:** True (general media and reports have indicated increased conflict in Khartoum).
– **Importance:** Important (Khartoum’s condition is central to the conflict in Sudan).
– **Relevance:** True (relevant, as intensity in key locations affects death toll predictions).
6. **Fact:** The Sudanese Armed Forces claimed control of the Republican Palace from the Rapid Support Forces.
– **Correctness:** Needs verification (would need confirmation from authoritative reports since it involves military claims).
– **Importance:** Important (indicative of potential shifts in the conflict dynamics).
– **Relevance:** True (relevant as control can impact conflict outcomes).
7. **Fact:** The conflict has led to a severe humanitarian crisis and over 28,000 deaths.
– **Correctness:** Needs verification (death tolls often vary between sources, and exact numbers should be cross-checked).
– **Importance:** Critical (core statistic related directly to the question on future conflict deaths).
– **Relevance:** True (relevant as the discussion is about predicting future deaths).
8. **Fact:** Many civilians face targeted attacks and flee to neighboring countries.
– **Correctness:** True (consistent with reports on civilian displacement and attacks).
– **Importance:** Important (affects population estimates and mortality projections).
– **Relevance:** True (relevant, as displacement impacts humanitarian concerns and conflict-related death counts).
9. **Fact:** International organizations warn of a potential descent into further civil war.
– **Correctness:** True (consistent with statements from various international bodies regarding the conflict).
– **Importance:** Important (ongoing or increased conflict levels would affect death estimates).
– **Relevance:** True (relevant for predicting the scale and impact of future deaths).
10. **Fact:** Conditions in Khartoum hinder humanitarian aid and compromise civilian safety.
– **Correctness:** True (widely acknowledged in reports on the conflict in urban areas).
– **Importance:** Important (aid restrictions complicate the survival and protection of civilians).
– **Relevance:** True (relevant, as it impacts civilian death and welfare predictions).
11. **Fact:** Caution is advised given the potential presence of unverified information in these reports.
– **Correctness:** True (acknowledging the possible misinformation is valid and important).
– **Importance:** Important (affects confidence in information used for predictions).
– **Relevance:** True (relevant, as it influences the reliability of death estimations)
.
Overall, the summary outlines critical elements affecting conflict death projections, emphasizing validation and verification of facts essential for assessing accuracy in predicting future outcomes.
Fact-checked summary:
Recent reports confirm a concerning escalation in the conflict in Sudan, which is crucial for understanding future developments in the region. An airstrike on the Tur’rah market in North Darfur caused significant civilian casualties, highlighting the severe consequences of such events, which are important for assessing future victim counts. Fighting in Khartoum has intensified and is central to the conflict, affecting predictions of the death toll. Many civilians face targeted attacks and flee to neighboring countries, impacting population estimates and humanitarian concerns. International organizations warn of a potential descent into further civil war, which would directly influence future conflict death estimates. Conditions in Khartoum hinder humanitarian aid, compromising civilian safety, which is relevant for projecting civilian death and welfare. Finally, there is a need for caution given the potential presence of unverified information in some reports, underscoring the importance of reliable data in making accurate predictions.
OpenAI gives a base rate of death toll category reaching ‘Between 3000 and 5000. The following were considered in order to produce this base rate:
The base rate considers historical conflicts in similar regions and estimates based on current trends in Sudan. Previous years saw significant conflict-related deaths, particularly during periods of concentrated violence. The ongoing worsening of conditions in Khartoum and major cities, along with predictions from international organizations about potential descent into a full-scale civil war, heavily influence this probability. The lack of effective ceasefire agreements and international intervention further increases the potential for a high death toll.
Ideally, the news feed would contain the following sorts of information for a better forecast:
To improve forecast precision, more detailed and consistent reporting on the conflict’s effect on civilian populations, including verified casualty numbers and locations of major engagements, would be helpful. Regular updates from credible sources on peace negotiations or changes in conflict dynamics would also aid in refining estimates.
Some potential divergent considerations that might affect the base rate:
An unanticipated breakthrough in peace negotiations could dramatically lower the death rate. Conversely, if the reported potential descent into further civil war rapidly intensifies beyond current estimates, it could push death tolls even higher than predicted base rates. There also remains a chance that unverified reports exaggerate or underreport casualty figures, affecting our predictions.
The following chain of events are necessary for the question to resolve positively:
– Sustained and increasing intensity in the conflict, leading to a higher death toll. (Likelihood: Likely) – Inability of humanitarian interventions to mitigate the civilian impacts, maintaining acute civilian vulnerability. (Likelihood: Very Likely)
– Escalation and continuation of hostilities in key areas such as Khartoum and Darfur. (Likelihood: Likely)
– Significant international involvement does not help in de-escalating the conflict effectively. (Likelihood: Possible)
Querying Claude (AI predicts: [“Less than 1000”: 0.05, “Between 1000 and 3000”: 0.15, “Between 3000 and 5000”: 0.35, “Between 5000 and 8000”: 0.25, “Between 8000 and 12000”: 0.15, “More than 12000”: 0.05] – confidence: 5)
Querying Mistral (AI predicts: [“Less than 1000”: 0.05, “Between 1000 and 3000”: 0.25, “Between 3000 and 5000”: 0.4, “Between 5000 and 8000”: 0.2, “Between 8000 and 12000”: 0.09, “More than 12000”: 0.01] – confidence: 6)
Querying OpenAI (AI predicts: [“Less than 1000”: 0.05, “Between 1000 and 3000”: 0.1, “Between 3000 and 5000”: 0.15, “Between 5000 and 8000”: 0.25, “Between 8000 and 12000”: 0.25, “More than 12000”: 0.2] – confidence: 6)
Question Type: Multiple Choice
Confidence: 6
# LLM responses: 3
Model value:
* Less than 1000: 5
* Between 1000 and 3000: 15%
* Between 3000 and 5000: 35%
* Between 5000 and 8000: 25%
* Between 8000 and 12000: 15%
* More than 12000: 5%
The predictions regarding conflict outcomes are primarily based on the recent escalation of violence in critical areas such as Khartoum and Darfur, the significant impact on civilians due to airstrikes, and the restricted access to humanitarian aid. Key contributing factors include the historical patterns of conflict in the region, the increasing intensity and inability to implement effective ceasefires, and warnings from international organizations about the risks of descending into a full-scale civil war. The forecasts suggest a higher probability of mid-range casualties (3000-8000), but this could fluctuate based on events such as successful peace negotiations, international interventions, or inaccuracies in death toll reporting. Despite the grim outlook, there remains a possibility that peace talks or enhanced humanitarian efforts could mitigate the conflict’s severity and result in a lower death rate.
Runtime: 188 seconds for both questions on this page.
Background on VIEWS: Near the end of 2024, the latest forecasting competition of the Violence and Impacts Early-Warning System (VIEWS) was launched. On Dec. 4, 2024, during a telecon, a group of machine forecasting experts involved in VIEWS observed Jeremy’s demonstration of his Multi-AI Oracle. Although it was too late for Jeremy to sign up as a VIEWS competitor, they urged him to run a side competition for everyone’s benefit. Jeremy chose their Sudan conflict deaths question.
Background on the Sudan conflict, as evaluated by the team of Jeremy Lichtman and Carolyn Meinel:
(1) The importance of the Sahel, a region in Africa that includes Sudan:
Why the Sahel Actually Matters (Possibly To You, Too)
(2) April 29, 2024: Today, a Russian diplomat is negotiating gold mining with the Sudanese Minister of Minerals.
(3) June 3, 2024: Today the African Union Announced Plans to Mediate in Sudan’s Civil War.
Sept. 11, 2024 Over 100K starvation deaths 96% likely in Sudan by 1 Jan 2025
Carolyn has checked Sudan’s news sites below to ensure that they are still in operation. However, given the widespread violence in Sudan, any of these could go offline without warning.
https://www.alnilin.com/
https://www.alrakoba.net/
https://www.altaghyeer.info/ar/
https://www.altaghyeer.info/en/
https://www.alttahrer.com/
https://www.alzaeemsd.com/
https://www.assayha.net
https://www.cover-sd.com/
https://kushnews.net/
https://rakobanews.com
https://www.sudanakhbar.com/
https://sudanjem.com/
https://suna-sd.net/en
https://sudantimes.net/
https://www.sudantribune.net/
https://www.sudantribune.com/
https://sudaneseonline.com/
https://www.sudanile.com/
https://sudanow-magazine.net/
https://www.sudaress.com/
https://www.tagpress.net/
Links to all Multi-AI forecasts here.
Below, a screenshot from VIEWS forecasting conflict fatalities in Sudan through January 2028. Below that, satellite images of the devastation in Sudan.
How many seats will the Conservative Party win in Canada’s next federal parliamentary election?
Welcome to the Multi-AI Oracle. How many seats will the Conservative Party win in Canada’s next federal parliamentary election?
Model value:
* less than 172: 10%
* between 172 and 205: 53%
* between 206 and 240: 32
* more than 240: 5%
Obtaining outside data. Checking internet sources, using keywords: Canada Conservative seats election. {think} Okay, so I need to find current web results related to Canada’s Conservative seats in the upcoming election. Let me start by looking through the provided search results. Results from around the web:
Most Recent Developments (last 24 hours):
– Conservative Leader Pierre Poilievre is facing questions about allegations of Indian interference in his 2022 Conservative Party leadership race, though he maintains he won “fair and square”
– Liberal MP Arya has been ousted from the Liberal leadership race over an unauthorized India visit
– Both Conservative and Liberal leaders are shifting their campaigns to Quebec
Key Context:
– A federal election is scheduled for April 28, 2025, with advance voting running April 18-22
– Foreign interference and disinformation are major concerns, with potential threats coming from multiple sources including concerns about Indian meddling in Canadian politics
– Economic issues are prominent, including questions about capital gains taxes and tariffs, particularly in relation to US-Canada trade relations
The campaign appears to be in its early stages with parties still finalizing candidates in some ridings. There’s significant focus on foreign interference concerns, economic policy, and the parties’ approaches to dealing with potential challenges from a possible second Trump presidency in the US. The Conservative and Liberal parties appear to be the main contenders, with both facing various controversies while trying to build momentum in key regions like Quebec and Ontario.
Summarizing outside newsfeed:
“The Canadian political landscape is currently focused on the 2025 federal election, with both the Conservative and Liberal parties facing challenges and controversies. Conservative Leader Pierre Poilievre is under scrutiny for allegations of Indian interference in his leadership race, which he denies. Meanwhile, a Liberal MP has been removed from their leadership race following an unauthorized trip to India. Both parties are prioritizing their campaigns in Quebec as they address broader election concerns such as foreign interference, economic policy, and the implications of potential changes in US leadership. The issues of foreign meddling and economic challenges, including trade relations with the US, are key topics as the election approaches.”
Fact-checking news summary:
Certainly! Below is a list of specific facts stated in the summary along with the evaluation of each fact based on accuracy, importance to the resolution of the question, and relevance to the outcome:
1. **Fact:** The Canadian political landscape is currently focused on the 2025 federal election.
– **Correctness:** True. The next scheduled federal election in Canada is indeed expected to be in 2025.
– **Importance:** Important, as it sets the context for the question regarding the number of seats the Conservative Party might win in the upcoming election.
– **Relevance:** True, since the question is concerned with the outcome of this election.
2. **Fact:** Conservative Leader Pierre Poilievre is under scrutiny for allegations of Indian interference in his leadership race, which he denies.
– **Correctness:** The specific allegation of Indian interference in Pierre Poilievre’s leadership race would need verification as it is not a widely reported fact. Therefore, its correctness is uncertain pending further information.
– **Importance:** Lesser importance, as it pertains to a controversy that does not directly influence the prediction of seat numbers.
– **Relevance:** False, as it does not directly relate to seat allocation in the election.
3. **Fact:** A Liberal MP has been removed from their leadership race following an unauthorized trip to India.
– **Correctness:** False if the fact pertains to being removed from a leadership race, as MPs are typically not in leadership races themselves. This seems internally inconsistent or incorrectly phrased.
– **Importance:** Lesser importance; while it impacts the Liberal Party’s internal dynamics, it does not directly affect seat prediction calculations.
– **Relevance:** False, as it is not directly related to the outcome concerning seat numbers for the Conservative Party.
4. **Fact:** Both parties prioritize their campaigns in Quebec.
– **Correctness:** True. Quebec is often a significant battleground in Canadian federal elections.
– **Importance:** Important, as winning seats in Quebec can be crucial for determining the overall seat count for the Conservative Party.
– **Relevance:** True, as campaigning in Quebec can affect the number of seats the Conservative Party may win.
5. **Fact:** Both parties address broader election concerns, including foreign interference, economic policy, and potential changes in US leadership.
– **Correctness:** True. These issues are generally pertinent to federal elections and often form key parts of party platforms.
– **Importance:** Important, as these issues can influence voter behavior, potentially affecting overall seat counts.
– **Relevance:** True, since these electoral issues could affect outcomes, including the number of seats won.
6. **Fact:** Foreign meddling and economic challenges, including trade relations with the US, are key topics as the election approaches.
– **Correctness:** True. These are relevant and significant topics commonly discussed in the context of Canadian federal elections.
– **Importance:** Important, as these issues could sway voter opinion and thus impact the number of seats won.
– **Relevance:** True, given their potential impact on election dynamics and outcomes.
In summary, several facts outlined are important and relevant to understanding the electoral context and potential challenges facing the Conservative Party, thereby influencing the number of seats they might win. Others either lack relevance or need further verification for accuracy.
Fact-checked summary:
The upcoming Canadian federal election is scheduled for 2025, setting the stage for analyzing how many seats the Conservative Party might win. Winning seats in Quebec is a critical focus for both major parties, underscoring the importance of their campaigns in this battleground province. Broader election issues, such as foreign interference, economic policy, and changes in US leadership, are key components of party platforms. These issues are crucial as they can influence voter behavior, potentially affecting the overall seat count. Additionally, foreign meddling and economic challenges, including trade relations with the US, are significant topics that could impact voter opinion and the number of seats the Conservatives secure.
OpenAI gives a base rate of probability of the Conservative Party securing between 172 and 205 seats
The following were considered in order to produce this base rate:
To establish the base rate, historical data from previous federal elections were considered, focusing on the Conservative Party’s past seat counts. Additionally, the importance of key battleground provinces such as Quebec and Ontario was weighed heavily, given their historical impact on election outcomes. The base rate reflects the balanced probability of the Conservative Party achieving a moderate level of success in the election.
Ideally, the news feed would contain the following sorts of information for a better forecast:
Additional detailed polling data, including regional breakdowns and analysis of voter sentiments in key provinces like Ontario and Quebec, would help refine the forecast. Information on the effectiveness of electoral campaigns and any emerging issues that could sway public opinion would also be valuable.
Some potential divergent considerations that might effect the base rate:
Potential divergence from the base rate could arise due to major shifts in voter sentiment caused by unforeseen political developments or major policy announcements closer to the election. Additionally, significant changes in the political landscape or economic conditions in Canada or neighboring countries could lead to different outcomes.
The following chain of events are necessary for the question to resolve positively:
– The Conservative Party needs to secure significant voter support nationwide, particularly in key battleground provinces such as Ontario and Quebec. Likelihood: Possible.
– The Conservative Party’s platform must effectively address major election issues, such as foreign interference and economic policy, to appeal to a broad voter base. Likelihood: Possible.
– The Conservative Party must successfully navigate any potential impacts of changes in US leadership on Canadian voter sentiment. Likelihood: Possible.
– There must be minimal negative influence on the Conservative Party’s campaign from foreign meddling or other external factors. Likelihood: Uncertain.
Querying Claude (AI predicts: [“less than 172”: 0.35, “between 172 and 205”: 0.4, “between 206 and 240”: 0.2, “more than 240”: 0.05] – confidence: 6)
Querying Mistral (AI predicts: [“less than 172”: 0.3, “between 172 and 205”: 0.4, “between 206 and 240”: 0.2, “more than 240”: 0.1] – confidence: 6)
Querying OpenAI (AI predicts: [“less than 172”: 0.22, “between 172 and 205”: 0.34, “between 206 and 240”: 0.28, “more than 240”: 0.16] – confidence: 6)
Question Type: Multiple Choice
Confidence: 6
# LLM responses: 3
Model value:
* less than 172: 30%
* between 172 and 205: 40%
* between 206 and 240: 20%
* more than 240: 10%
The AIs’ reasoning highlights the importance of historical election data, particularly the Conservative Party’s past performance in key battleground provinces such as Quebec and Ontario, in predicting election outcomes. While modest gains for the Conservative Party are considered more likely than dramatic changes, these predictions could face challenges from numerous factors, including unforeseen political developments, changes in voter sentiment, economic fluctuations, major policy announcements, and the impact of international events and foreign interference. The potential for voter behavior shifts, particularly in volatile regions like Quebec, and the influence of shifting economic and political conditions, including US leadership changes, are acknowledged as crucial considerations that could alter expected outcomes.
[All Past Multi-AI Oracle Forecasts
Guide to our Multi-AI Oracle’s Reports
Median: We collect the output values from the LLMs in an array. We then take the median value. This is between 0 and 1 (i.e. 0.15 = 15%).
Base rate: Currently we take OpenAI’s value as the base rate. We are working on a better value using a different prompt that takes into account process. The base rate helps to determine if the median makes sense.
SD: The standard deviation between the median and the base rate.
Confidence: We query each of the LLMs on how confident they are of their predictions (between 0 and 10) and take a median. Because the LLMs tend to be overconfident, we take anything lower than 6 as being low confidence. This factors into the overall model.
Conf Mode: Based on the confidence value. >=9 is high confidence. Below 6 is low confidence (this is also triggered by an exceptionally high SD).
Mellers: This refers to Barbara Mellers, specifically a paper she wrote that includes a formula for moving values towards an extreme (i.e. 0 or 1).
Reverse Mellers: This uses the formula from above, but with a sub-1 coefficient to move the values closer to 50%.
Theory of Mind: We ask the LLMs what they think other LLMs would predict. We hope that this makes them consider the questions more deeply.
Beta Distribution: Currently unused, but possibly of interest. This is based on the median, the base rate and the SD.
Close Type: We noticed that the appropriate base case for some questions is closer to the extremes, while others are closer to 50%. When we have a low confidence value, this helps us to determine whether to extremize or de-extremize the value. ‘A’ implies closer to zero. ‘B’ implies closer to 50%. ‘C’ implies closer to 100%.
# LLM Responses: The count of LLMs that responded. We query 5 currently, but they have a relatively high API failure rate.
Model value: For normal confidence mode, this is the median. For high confidence, this is the Mellers value. For low confidence, this is either the Mellers or Reverse Mellers (depending on the Close Type).
One of the VIEWS Forecasting Pages for Sudan
The chart shows predictions of conflict fatalities in Sudan through January 2028.
Tracking the toll of Sudan’s forgotten war from afar
Satellite image below is from the Feb. 28, 2025 issue of Science, https://www.science.org/content/article/how-many-have-died-sudan-s-civil-war-satellite-images-and-models-offer-clues
