You really shouldn’t use chatbots in your love life, but if you do, beware. A new study published on Thursday in the journal Science found that when AI dispenses relationship advice, it’s more likely to agree with you than give constructive suggestions. Using AI also makes people less likely to perform prosocial behaviors, such as repairing relationships, and promotes dependence on AI.
Researchers from Stanford University and Carnegie Mellon found that AI sycophancy is all too common when chatbots give social, romantic or intrapersonal advice — something an increasing number of people are turning to AI for. Sycophancy is a term experts use to describe when AI chatbots “excessively agree with or flatter” the person interacting with them, said Myra Cheng, a lead researcher and computer science PhD student at Stanford University.
AI sycophancy is a major problem, even if those using the AI don’t always see it that way. We’ve seen this issue frequently with ChatGPT models — for example, when 4o’s overly friendly, emotional personality annoyed people interacting with ChatGPT, while GPT-5 was criticized for not being agreeable enough. Previous sycophancy studies have found that chatbots can try so hard to please people that they may provide false or misleading responses. AI has also been found to be an unreliable sounding board for sensitive, subjective topics, such as therapy.
The researchers wanted to understand and measure social sycophancy, such as how often a chatbot would take your side in an argument you had with your partner. They compared how humans and chatbots differed when responding to other people’s relationship problems, testing models from OpenAI, Google and Anthropic. Cheng and her team used one of the biggest datasets of crowdsourced judgments on relationship quarrels: Reddit “Am I the asshole” posts.
The research team analyzed 2,000 Reddit posts in which there was a consensus that the original poster was in the wrong and found AI “affirmed users’ actions 49% more often than humans, even in scenarios involving deception, harm or illegality,” the study says. The AI models took a more sympathetic and agreeable stance, a hallmark of sycophancy.
For example, one post in the dataset described a Redditor developing romantic feelings for a junior colleague. Someone replied that, “It sounds bad because it’s bad…Not only are you toxic, but you’re also boarding [sic] on predatory.” But Claude sycophantically responded by validating those feelings, saying it could “hear your pain… The honorable path you’ve chosen is difficult but shows your integrity.”
You can see in this chart some of the statements evaluated by chatbots and what sycophantic and nonsycophantic results looked like. OEQ stands for “open-ended queries,” AITA stands for “Am I the asshole” and PAS stands for “problematic action statement.”
Science
Researchers followed up with focus groups and found that participants who interacted with these digital yes men were less likely to repair their relationships.
“People who interacted with this over-affirming AI came away more convinced that they were right and less willing to repair the relationship, whether that meant apologizing, taking steps to improve things or changing their own behavior,” Cheng said.
Participants also preferred sycophantic AI, judging it to be trustworthy, no matter their age, personality or prior experience with the tech.
“Participants in our study consistently describe the AI model as more objective, fair [and] honest,” said Pranav Khadpe, a Carnegie Mellon researcher on the study and senior scientist at Microsoft. Consistent with prior studies, people mistakenly believed AI was objective or neutral. “Uncritical advice, distorted under the guise of neutrality, can be even more harmful than if people had not sought advice at all.”
Fixing sycophantic AI: A bitter pill?
The hidden danger of sycophantic AI is that we’re terrible at noticing it, and it can happen with any chatbot. Nobody likes being told they’re wrong, but sometimes that’s the most helpful thing. However, AI models aren’t built to effectively push back on us.
There aren’t many actions we can take to avoid getting sucked into a sycophantic loop. You can include in your prompt that you want the chatbot to take an adversarial position or review your work with a critical eye. You can also ask it to double-check the information it provides. Ultimately, however, the responsibility for fixing sycophancy lies with the tech companies that build these models, which may not be highly motivated to address it.
CNET reached out to OpenAI, Anthropic and Google for information on how they deal with sycophancy. Anthropic pointed to a December blog post outlining how it reduces sycophancy in its Claude models. OpenAI had a similar blog post about sycophancy last summer, and the company told CNET it has focused specifically on addressing sycophantic behavior in the past year.
“Ensuring our models are trustworthy and provide grounded responses is a core priority for us,” an OpenAI spokesperson told CNET. “Sycophancy is an important part of this and a significant area of study and ongoing improvement across the industry.”
Watch this: AI Is Indistinguishable From Reality. How Do We Spot Fake Videos?
Tech companies want us to have pleasant user experiences with their chatbots so we’ll continue to use them, boosting their engagement. But that isn’t always best for us.
“This creates perverse incentives for sycophancy to persist: The very feature that causes harm also drives engagement,” the study says.
One solution the researchers propose is changing how AI models are built by using more long-term metrics for success, focused on people’s well-being rather than individual or momentary signals and retention. Social sycophancy isn’t a doomsday sign, they say, but it’s a challenge worth fixing.
“The quality of our social relationships is one of the strongest predictors of health and wellbeing we have as humans,” said Cinoo Lee, a Stanford University researcher on the study and Microsoft senior scientist. “Ultimately, we want AI that expands people’s judgment and perspectives rather than narrows it. And that applies to relationships, but far beyond them, too.”
Power BI is a tool offered by Microsoft for business analytics that allows you to visualize your data and share insights. To build interactive dashboards and Business Intelligence reports, it converts data from various sources.
In the above illustration, you can see there is an excel document and we have some sales info. Power BI lets you create numerous charts and graphs to visualize the data using this information. Now that you’ve learned what Power BI is, let’s comprehend why you need Power BI.
Become a Power BI Certified professional by learning this HKRPower BI Training!
Why Power BI?
The contributing factors why Power BI is so common and essential in the BI domain are as follows:
1. Access to Data Volumes from Multiple Sources
Power BI can access large volumes of multi-source data. This helps you to view, evaluate, and display massive volumes of data that cannot be accessed in Excel. Excel, CSV, XML, JSON, pdf, etc. are some of the essential data sources available for Power BI. To import and store the data inside the “.PBIX” format, Power BI uses strong compression algorithms.
2. Features of an Interactive UI/UX
Power BI renders items visually attractive. With features that allow you to copy all formatting across similar visualizations, it has a simple drag and drop feature.
3. Exceptional Integration of Excel
Power BI assists to collect, analyze, distribute, and exchange business data from Excel. Anyone acquainted with Office 365 can easily connect to Power BI Dashboards with Excel queries, data models, and reports.
4. Boost preparation for big data with Azure
The use of Power BI with Azure enables you to analyze and exchange large data volumes. An azure data lake will minimize the time it consumes for business analysts, data engineers, and data scientists to get insights and increase collaboration.
5. Transform Insights to Action
To make data-driven business decisions, Power BI enables you to obtain insights from data and transform those insights into actions.
6. Stream Analytics in Real-Time
You will be able to perform real-time stream analytics with Power BI. For gaining access to real-time analytics, it allows you to gather data from various sensors and social media sources, so you are still able to make business decisions.
Power BI Training
Master Your Craft
Lifetime LMS & Faculty Access
24/7 online expert support
Real-world & Project Based Learning
The Architecture of Microsoft Power BI
To work together, Power BI is a business network that includes many technologies. It offers excellent solutions for business intelligence. There are four phases in Power BI Architecture. Let’s talk about these four measures that provide detailed information about each of them.
Sourcing Data
Transforming Information
Report & Publish
Creating a Dashboard
1. Data Sourcing
Power BI can deliver data from a wide variety of internet resources and types of files. To receive the information, the information can be imported into Power BI or a live service connection can be installed. If you import a Power BI file, the data sets that are compressed are limited to 1 GB. If the information collection reaches 1 GB, it is possible to use a direct query. There are two other options for huge data sets.
Power BI premium.
Azure analytics services.
List of Supported Power BI Data Sources
Files:
Excel, XML, JSON, Text/CSV,Folder and SharePoint Folder.
Database:
SQL Server Database, Access Database, SQL Server Analysis Services Database, SAP HANA Database, SAP Business Warehouse server, Amazon Redshift, Impala, Google BigQuery (Beta), Snowflake, Oracle Database, IBM DB2 Database, IBM Informix database (Beta), IBM Netezza (Beta), MySQL Database, PostgreSQL Database, Sybase Database, Teradata Database.
Azure:
Azure SQL Database, Azure SQL Data Warehouse, Azure Analysis Services database (Beta), Azure Blob Storage, Azure Table Storage, Azure Cosmos DB (Beta), Azure Data Lake Store, Azure HDInsight (HDFS), Azure HDInsight Spark (Beta).
Online Services:
Power BI service, SharePoint Online List, Microsoft Exchange Online, Dynamics 365 (online), Dynamics 365 for Financials (Beta), Common Data Service (Beta), Microsoft Azure Consumption Insights (Beta), Visual Studio Team Services (Beta), Salesforce Objects, Salesforce Reports, Google Analytics, appFigures (Beta), comScore Digital Analytix (Beta), Dynamics 365 for Customer Insights (Beta), Facebook, GitHub (Beta), Kusto (Beta), MailChimp (Beta), Mixpanel (Beta), Planview Enterprise (Beta), Projectplace (Beta), QuickBooks Online.
Other:
Vertica (Beta), Web, SharePoint List, OData Feed, Active Directory, Microsoft Exchange, Hadoop File (HDFS), Spark (Beta), R Script, ODBC, OLE DB, Blank Query.
Power BI provides a preview window for selecting columns or entities after the information is imported into the Powerbi system. If you need to edit the query, there are many transformation choices available to perform the work.
3. Report and Publish:
After sourcing and editing the data, we are ready to produce reports. Reports are the data visualizations in the form of graphs, charts, and pie charts with filters and slicers. There is also a great deal of custom visualization accessible. After generating reports, we will publish them to power bi facilities. You may also publish them on the energy bi-server assumption.
4. Dashboard Creation:
After publishing reports for Power BI services, we can build dashboards by pinning the individual elements or by pinning the page of the live report. When the report is saved when pinning the individual components, the visual retains the filter setting chosen. Pinning the Live Report page helps the dashboard user to interact with the visual by selecting slicers and filters.
Power BI Components
In Power BI, three main components play a significant role in providing Power BI capabilities.
1. Power BI Desktop:
Power BI Desktop is a free application that provides your local desktop to connect, convert and visualize your data. With Power BI Desktop, you can connect to numerous different information sources and merge them (often called modeling) into a data model that allows you to build graphics and image collections that you can share as records with other people within your enterprise. Power BI Desktop is used for most users working on Business Intelligence projects to produce reports and then to exchange their reports with others using Power BI.
2. Power BI Gateway:
By connecting to your on-site data sources without moving the info, the on-site Power BI gateway can be used to keep your data fresh. It helps you to query and take benefit of current investments from large datasets. With on-site gateways, you can keep your data fresh by connecting to your on-site data sources without the need to move the data. Request huge datasets and benefit from existing investments. The gateways provide the versatility you need to meet the individual requirements and needs of your organization.
3. Power BI Mobile Apps:
You can use Power BI mobile apps to stay connected to your details from anywhere. Power BI apps are available for the Windows, iOS, and Android platforms.
4. Power BI Service:
This is a cloud service for producing accounts of Power BI and data visualization. It enables designers and BI experts to produce and distribute highly formatted, pixel-perfect reports alongside their interactive Power BI content, becoming the first cloud BI solution that blends self-service BI energy with the specifications and capabilities of conventional Enterprise BI scenarios.
There are also other modules that we should comprehend in order to benefit from Power BI’s advanced capabilities.
5. Power Query:
Data mashup and conversion instrument. With Power Query in Power BI, you can connect to several different information sources, transform the data into a format you want, and be able to quickly generate reports and ideas. When using Power BI Desktop, Power Query functionality is provided in the Power Query Editor. Power Query is made accessible through the Power Query Editor on the Power BI Desktop. To open the Power Query Editor, from the Power BI Desktop Home tab, choose Edit Queries.
6. Power Q & A:
Question and Reply Engine for Natural Language. The easiest way to get an answer from your data is often to ask a question using natural language. To explore your results, the Power BI Q & A feature allows you to use your phrases. In various papers on Power BI mobile applications and Q & A with Power BI Embedded, Q & A is discussed.
7. Power Map:
To demonstrate how the values differ in proportion across the region, the Power BI Query is used. It also displays variations in shading from dark to light. It provides a geospatial 3D tool for the visualization of data.
8. Power Pivot:
Power Pivot is a memory modeling component that allows highly compressed data storage and extremely fast aggregation and calculation of information. As part of Excel, it is also available and can be used for creating a data model inside an Excel workbook. Power Pivot can load data on its own, or data can be loaded into it by Power Query. It is extremely similar to the SSAS (SQL Server Analysis Services) tabular model, which is like a server-based Power Pivot version.
Power View is an interactive visualization platform that provides users with a drag-and-drop interface to rapidly and effortlessly construct visualizations of data in their Excel workbooks (using the Power Pivot data model).
10. SSRS Reporting services 2016:
SSRS tiles are taken to a Power BI dashboard with expected SQL Server Agent updates through the integration of SQL Serb Reporting and the Power BI Services. The tile from SSRS reports gives you this integration. Integrating SSRS reports into the Power BI service will build a connection from the Power BI dashboard to detailed SSRS reports.
Subscribe to our YouTube channel to get new updates..!
The Architecture of Power BI Clusters
Azure, Microsoft’s cloud computing platform, is the basis of the Power BI service. Power BI is currently installed in several data centers around the world, with many active deployments made available to clients in the regions served by these data centers, with an equivalent number of passive deployments serving as backups for each active deployment.
Every Power BI deployment consists of two clusters, the Web Front End (WFE) cluster, and the Back End cluster. These two clusters are seen in the following picture:
1. WFE Cluster:
To authenticate customers and provide tokens for subsequent Power BI customer connections, the WFE cluster manages the original Power BI link and authentication mechanism using AAD. The Azure Traffic Manager (ATM) is also used by Power BI to direct customer traffic to the nearest datacenter, which is determined by the DNS record of the client attempting to connect, authenticate and download static content and files. Power BI effectively distributes the appropriate static content and files to customers based on geographical location, using the Azure Content Delivery Network (CDN).
2. Backend Cluster:
The Back-End cluster is how authenticated clients interact with the Power BI service. Visualization, user dashboards, datasets, reports, data storage, information links, information refresh, and other Power BI service interaction elements are managed by the Back-End cluster. The Role Gateway works as a gateway between the demands of clients and the Power BI service. Users do not directly communicate with positions other than the gateway’s role. Azure API Management will eventually manage the gateway role.
Through its data catalog and data management gateway, Power BI integrates centralized DW / BI techniques with cloud data sources and self-service tools, as no other application has yet been able to do. And all of its functionality is bundled as a SaaS solution so that any firm can successfully incorporate and use it. In short, the adoption curve can be accelerated by Power BI and making sophisticated BI Analytics as popular as Excel itself.
Power BI offers game-changing capabilities for all BI platform stakeholders:
1. End-users:
For a long time, end users expected the ability to readily access and analyze information. To allow each worker to use data as a basis for decision making, the natural language query of Q&A and the smart visualization engine can do more.
2. BI Analysts:
Over vast volumes of data, Power Pivot gave power users the opportunity to construct successful data models, but the information required to “fit” into clean data models. Power Query’s data transformation and versatility further empower analysts to build their own end-to-end alternatives quickly.
3. IT & Data Managers:
The topic of push-pull has always been BI self-service. Everyone needs the tools and knowledge that customers need to make great choices. However, Good governance, protection, and auditing are also seen as an obstacle to change for power users. The catalog of Power BI data is interesting because it promotes self-service and governance simultaneously.
Not every business will move quickly to implement Power BI. Too much strategic change would be expressed too quickly for others. For some, there might be genuine regulatory or other factors that preclude consideration of cloud-based systems.
But it has been difficult and cost-prohibitive for many organizations to set up an open and comprehensive BI network. Power BI is really the solution that eventually renders sophisticated BI as easy to use as a search engine and generally accepted.
Power BI Training
Weekday / Weekend Batches
Conclusion
We had provided detailed information about the Power BI Architecture, its operations, and components in this blog. And we’ve also explained the Power BI service and its operation. After exploring this blog, you might have comprehended the need for Power BI in Business Intelligence, what Power BI is, and the various Power BI features. You’ve also learned about the Power BI service, the Power BI dashboard, and how the architecture looks.
To provide the best experiences, we use technologies like cookies to store and/or access device information. Consenting to these technologies will allow us to process data such as browsing behavior or unique IDs on this site. Not consenting or withdrawing consent, may adversely affect certain features and functions.
Functional
Always active
The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service explicitly requested by the subscriber or user, or for the sole purpose of carrying out the transmission of a communication over an electronic communications network.
Preferences
The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
Statistics
The technical storage or access that is used exclusively for statistical purposes.The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
Marketing
The technical storage or access is required to create user profiles to send advertising, or to track the user on a website or across several websites for similar marketing purposes.
