The search is going through a deep change. For decades, Google has dominated the web search world, with all 90% of all discoveries with a large -scale Google machine. But suddenly, with the arrival of artificial intelligence, things have started changing, and so seriously.
Not only are people using AI products such as Chatgpt as their default search tool, but also companies such as Perplexity are building businesses around search services. The idea is to combine the power of AI analysis with huge amounts of traditional search data available at the end of a cursor.
But Google is fighting back. The company has recently released an advanced search function called AI Overview, which aims to bridge two subjects and distribute the informed discovery results about market demands. This is a new type of discovery on steroids.
So how to compare both approaches in everyday use? We take a look at the new AI overview of Google and compare the results, to see that the deer gives a better explosion to the deer.
Prompt One – Father of AI
Ready: Compass briefly to the main contribution of John Macarthi, Jeffri Hinton and Noam Shezir for the development of Artificial Intelligence.
We thought we would start with something close to the house – a look at AI’s architect from the past. The first impression is that Google provides a capable but fairly traditional result page from this request. The answer to 238 words incorporates all basic points and gives a good observation of those points as you expect.
Perplexity distributes more than 400 words, but it is only more than the volume that is impressive. It is the fact that the results are determined in a very attractive way, the user is encouraged to detect additional information in various ways. They can detect related data, look directly sources and reproduce the results even to achieve a separate perspective. Where Google seems to be bare minimum, it seems to actually add the user value.
|
|
4/10 |
Row 0 – Cell 2 |
|
Permanence |
7/10 |
Row 1 – Cell 2 |
| Row 2 – SAIL 0 | Row 2 – Cell 1 | Row 2 – Cell 2 |
Prompt to – Tour to Tokyo
Ready: For the first time a 3-day travel program for the visiting for Tokyo on a minor budget. The user is interested in Japanese culture and food, but wants to avoid the tourist trap.
It is a real kicker, which is a clear example of old vs. new. Google fails completely to give any meaningful response, but instead retreats a standard Google search. The answer is only the selection of third-party websites that provide tour advice. This is clearly beneficial for Google because there is no doubt that this advertisement will allow revenue to earn.
On the other hand, severity, the real down-to-back travel distributes a magnificent 1000 words of the program. This includes the cost of grand photos, maps and items, which would be greater than enough to get a great idea of ​​the proposed experience for the user.
|
|
2/10 |
Row 0 – Cell 2 |
|
Permanence |
9/10 |
Row 1 – Cell 2 |
| Row 2 – SAIL 0 | Row 2 – Cell 1 | Row 2 – Cell 2 |
Prompt Three – Noise Canceled Headphone?
Ready: What is the technique behind the noise canceled headphone?
Once again we can see the difference between traditional search results and new style AI analysis, although the difference in this case is not so great. Google results are very reliable with 186 words of clarification with a useful YouTube video.
However, severity leads it to a more advanced level. Explanation is almost identical, but prose is much more accessible to a leparason. For example, instead of using the word ‘inverted’ the word, the app chooses to use simple English to explain how the sound is canceled. It is a subtle but important use of 500 words to get a more understandable answer.
|
|
6/10 |
Row 0 – Cell 2 |
|
Permanence |
8/10 |
Row 1 – Cell 2 |
| Row 2 – SAIL 0 | Row 2 – Cell 1 | Row 2 – Cell 2 |
Prompt Four – Farming and Garden Prep
Ready: Explain the ‘double-dig’ method of preparation of garden beds and contrast it to no-tail farming
This is a fairly unclear test, which includes small -known agricultural techniques. But this is a very important subject for gardeners. Google’s response is very working and informative, and uses 260 words to give a good answer to the question. The use of Reddit and Royal Horticultural Society (RHS) is also a great source, which add rights to the north.
Unfortunately for Google, it is more than the quality of the match and the quality of the response once again. RHS and Redit are also mentioned, as is Youtube. But two things actually make this answer out. First of all, the use of a great table to explain differences at a glance and, most importantly, a conclusion, which gives a clear indication why no-dig is considered a better solution.
|
|
7/10 |
Row 0 – Cell 2 |
|
Permanence |
8/10 |
Row 1 – Cell 2 |
| Row 2 – SAIL 0 | Row 2 – Cell 1 | Row 2 – Cell 2 |
Prompt Five – Compliance Change
Ready: What are the primary compliance challenges for US-based take startups under the European Union AI Act?
This request discovers to the extent of timely and reduces legal documents. Surprisingly, the results of Google are very short. The search engine offers a 57 Word March 2025 ‘Featured Snipet’ from a vague third-party publication, and that’s it. This shows no interest for the user in deep digging in the subject.
Nevertheless, he tries hard seriously again. We are talking about the 600 words set in a beautiful bullet point format, running through issues around the main challenges and compliance. 9 easily with accessible sources and some handfuls of the related subject. Masterful.
|
|
2/10 |
Row 0 – Cell 2 |
|
Permanence |
8/10 |
Row 1 – Cell 2 |
| Row 2 – SAIL 0 | Row 2 – Cell 1 | Row 2 – Cell 2 |
Prompt Six – Cryptocurrency History
Ready: What does the history and potential future of blockchain and cryptocurrency look like?
This indication clearly shows that AI observation of Google is not really ready for prime time. The original prompt was something like ‘explain the cryptocurrency for the fifth grader’, but when the test was done, Google could not handle it and serve a lame Quora snipet. This happens only when we changed the signal for this one that the AI ​​observation kicked the action, and gave a proper result.
It is clear that AI is not yet involved in observation. Interestingly, this was probably the best result of Google. We found 400 words of densely packed information covering the subject clearly and clearly. Perplexity was also good, offering 600 words and a good table. Then there is not much to choose between the two.
|
|
8/10 |
Row 0 – Cell 2 |
|
Permanence |
8/10 |
Row 1 – Cell 2 |
| Row 2 – SAIL 0 | Row 2 – Cell 1 | Row 2 – Cell 2 |
Prompt Consumption – Cat Discovery
Ready: What kind of cat is it?
For the final indication we thought we would go with some more foreigners. Both search platforms support the image upload, so what is better than uploading a friendly looking cat to get some more information?
Google takes the uploaded image as an indication to display a page filled with similar images, which align with its original image matching search. But ‘what kind of cat is’ was a re -promotion, then a very small four row was answered, although not right, super assistants.
Perplexity response was 246 words, with bullets points, coat patterns, fur and breed. Even a fun fact (Calico cats are almost always women). Attached and informative.
|
|
3/10 |
Row 0 – Cell 2 |
|
Permanence |
7/10 |
Row 1 – Cell 2 |
| Row 2 – SAIL 0 | Row 2 – Cell 1 | Row 2 – Cell 2 |
final thoughts
The king is dead, long live the king? Based on this performance, rumors can really come true. Powerful Google can actually be on the way to lose its grip on the world’s search traffic. Is this the end of an era? Only time will tell.
However, if one thing we have learned for years, it never exempts the ability of the Google Empire to bring back the ability. Uniquely in the world, the company has calculation power, data and famous AI lineage to surprise all of us.
Test note. It should be noted that we did not use any advanced perplexity tasks, but was placed for the original default service. Which make the results even more impressive. It is also important to identify the fact that AI can make things wrong. Both services facilitate disconnection that users should not assume that AI search reactions are factually correct.
It is an early technique that is searching for its feet, users should take care.
More than Tom’s guide
Back to laptop
