AI and neural networks for finding information online: 7 prompts for ChatGPT and Perplexity

7 best prompts for ChatGPT and Perplexity to use AI for finding information online

Searching for data online is something no modern person can do without in their daily life. However, it isn't always easy to phrase a query precisely and quickly find reliable, easy-to-digest information. In response to a query, search engines can return hundreds of links, but not every user can tell which of the offered data is outdated and which has no real basis at all.

This is where AI agents for finding information, as well as neural networks like ChatGPT and Perplexity, can help. They don't just extract keywords from a query — they understand its context, and they can analyze, compare, and explain complex concepts in plain language.

The GPTunnel neural network aggregator offers both services in a single interface. A free basic version is available for ChatGPT, while more advanced models require payment. Perplexity, on the other hand, is a paid tool from the start, billed on a pay-as-you-go basis.

How to phrase queries for neural network search correctly

A common mistake when working with AI is writing prompts the way you'd type a query into a regular search engine. AI for online information search "thinks" differently. For a neural network, what matters isn't matching words but the meaning of the task. To get a good result:

1. Phrase your query as an instruction. Not "Best startups," but "Make a table of 5 fast-growing AI startups from 2025 with a short description of their products." The more specific the question, the more precise and informative the answer.
2. Add a format (table, list, brief, detailed), and state the goal ("for a report," "for a post," "for analysis"). If you're interested in a comparison, write "compare"; if you need to verify a fact, write "check."
3. Don't be afraid to ask follow-up questions. AI models are built for dialogue. You can ask an additional question, request a more detailed explanation, ask for a shorter version, simplify it, or translate the text into another language.

7 best prompts for ChatGPT and Perplexity

Here are 7 templates to help you use neural networks as a professional research assistant. Each prompt can be adapted to any topic you're interested in and run in either ChatGPT or Perplexity.

Find precise data or facts

Prompt:

"Find the latest data (2024-2025) on global inflation, cite sources, and briefly explain what it means for the global economy."

ChatGPT result (GPT-5):

• global inflation (headline CPI) — around 5.9% in 2024 and 4.7-4.8% in 2025;
• core inflation (excluding food and energy) — around 4.5% and 3.7-3.8% respectively.

The model explained that disinflation is happening unevenly. Goods prices are falling faster than service prices, and wage growth in some countries is putting pressure on prices. The AI drew conclusions about investment activity and provided several links to sources.

You can review the full ChatGPT result via this link.

Perplexity result

Perplexity provided up-to-date data from international forecasts:

• Global inflation fell from 9.4% (2022 peak) to 3.4-3.5% in 2025;
• The IMF, UN, and World Bank agree: developed countries are returning to target levels (around 2%), while one in five developing economies is still facing double-digit inflation.

Perplexity also provided analysis and a large number of links to authoritative sources.

The neural network's results are reflected in this conversation.

Compare approaches or opinions on a single topic

Prompt:

"Compare how OpenAI and Google approach developing artificial intelligence and integrating AI services into their products."

ChatGPT result (GPT-5)

ChatGPT offered a comparison of OpenAI's and Google's strategies through the lens of 6 areas:

• Research;
• Infrastructure;
• Safety;
• Data;
• Integration;
• Monetization.

ChatGPT noted that OpenAI's competitive advantage is speed of innovation and UX, but it depends on Microsoft/Azure. Google, meanwhile, turns out to be more resilient thanks to its own infrastructure and a multibillion-strong base of regular users.

You can evaluate this example of the neural network's work in this chat.

Perplexity result

This model offered a compact but fact- and source-rich overview for each product. At the end of the overview there's a comparison table reflecting the contrast in approaches: OpenAI focuses on rapid development and experimentation, while Google's key trait is stability and independent infrastructure.

You can see how Perplexity performed here.

Create a topic overview

Prompt:

"Summarize the changes in online shopping in 2024-2025. Name 3-4 key trends (for example, marketplace growth, social commerce, mobile payments, sustainable consumption). Provide data and cite sources."

ChatGPT result (GPT-5)

GPT-5 conducted an analysis and identified 4 key trends in global e-commerce:

• Social media as a driver of online shopping;
• Generative AI built into the buyer journey;
• Growth of digital wallets in online payments;
• Same-day delivery and pickup becoming the new norm.

The neural network also concluded that 2025 would be the year of "smart retail," thanks to AI assistants, personalization, and fast logistics.

You can review the report by opening this chat.

Perplexity result

True to form, Perplexity focused on figures, facts, and sources, zooming in on a narrow slice of the market. The neural network noted the dominance of marketplaces in online shopping, growth in the number of orders, shrinking basket sizes, and the spread of online retail beyond major cities into smaller regions.

To evaluate the AI's work, follow this link.

Find specific examples or case studies

Prompt:

"Give 5 specific examples of how major brands implemented sustainable and eco-friendly solutions in 2024-2025. Describe which technologies or initiatives they used (for example, packaging recycling, emissions reduction, renewable energy), and cite sources."

ChatGPT result (GPT-5)

The neural network provided 5 examples of corporations that embraced the trend toward eco-friendly solutions:

• Microsoft;
• Google;
• Amazon;
• Starbucks;
• Maersk.

For each example, an official source is provided, and the analysis describes the technologies and real-world effect of implementing environmental protection measures.

You can check out GPT's overview here.

Perplexity result

Perplexity offered case studies from a specific regional market and, as always, focused on concise summaries and a large number of sources. Its brief overview mentioned several well-known retail chains and grocery brands from that market.

You can personally evaluate the neural network's work in the corresponding conversation.

Verify information or a rumor

Prompt:

"Check whether it's true that electric vehicles are more harmful to the environment than gasoline-powered cars. Provide arguments for and against, based on recent research (2024-2025), cite sources, and give a brief conclusion."

ChatGPT result (GPT-5)

GPT-5 refuted the claim made in the prompt. Drawing on a number of sources, the neural network compared the life cycle of electric vehicles and internal combustion engine cars and concluded that the former are 2-3 times cleaner in terms of CO₂e emissions than gasoline-powered vehicles.

To review the example, go here.

Perplexity result

Perplexity reached a similar conclusion but presented it in a brief, bullet-point overview. The summary is packed with figures and sources that can confirm the facts. The analysis is also presented in a "for/against" format.

You can find the neural network's report here.

Find reliable sources

Prompt:

"Find 10 reliable academic sources on the topic 'The development of Impressionism and its influence on contemporary art.' Cite authors, publication titles, publication years, and briefly describe how each source is useful for research (for example, historical context, style analysis, influence on contemporary painting or design)."

ChatGPT result (GPT-5)

GPT-5 selected 10 classic and contemporary academic sources — from John Rewald and T. J. Clark to museum catalogs and critical anthologies. Each source comes with a note on the value it can add.

You can view the selection here.

Perplexity result

The neural network offered 10 sources hosted mostly on popular platforms, all in a single shared language — which makes it easier to quickly grasp the context and assess how useful the information is.

To see the task result, follow this link.

Quickly get up to speed on a new topic

Prompt:

"Explain what string theory is in simple terms, as if for a beginner. Add examples, comparisons, and recent facts from 2024-2025 to show how this theory is developing and why physicists need it."

ChatGPT result (GPT-5)

GPT-5 explained string theory as "the orchestra of the universe," where particles are different notes of a single vibrating string. The neural network described the essence of the theory, explained why physicists need it, and provided data on the latest developments. However, despite the overview's light tone, the explanation contained a lot of complex terminology, which made some parts hard to follow.

You can judge how well GPT handled the task by viewing this conversation.

Perplexity result

Perplexity gave a precise but fairly simple explanation of the theory. Compared to GPT, Perplexity's explanation is much easier for someone unfamiliar with physics to understand.

And, as usual, the neural network provided a large number of additional sources.

You can check out its take on string theory via this link.

Prompt templates for ChatGPT and Perplexity

ChatGPT vs Perplexity

These tools are considered among the best AI options for finding information. However, there are differences in how they work. For example:

• ChatGPT is a source of detailed answers with reasoning and thorough explanations that follow clear logical connections. It's great when you need to quickly get up to speed on a topic, write an analytical note, or put together a short overview.
• Perplexity solves different tasks. Its strength is speed and accuracy. The neural network searches for data live and backs it up with links to sources, so any given fact can be quickly verified.

Simply put, ChatGPT is a conversation partner that helps you quickly understand a topic. Perplexity is a reliable assistant for fact-checking. To get an accurate and clear result when searching for information, it's best to combine the two tools. Start by searching for and verifying facts with Perplexity, then feed them to ChatGPT, which will analyze the data and organize everything clearly.

Conclusion

A neural network for finding and analyzing information is now a component no one can do without — whether in business or in education.

Models like GPT and Perplexity help you find, analyze, structure, and verify data. The key is to skillfully use the strengths of each tool and write correct prompts. For example, GPT can help put together a detailed analysis with thorough explanations, while Perplexity lets you quickly verify facts.

A smart combination of available neural networks makes it possible to cut the time spent searching for and analyzing large volumes of information from several days down to just a couple of hours.