Recently, Google Search added a feature called &#34;AI Overview&#34; that can directly display a short answer at the top of search results. For example, if you search &#34;how to clean leather boots,&#34; the system will immediately show a multi-step cleaning guide that integrates information from the web. This feature allows you to quickly obtain information without having to browse through multiple web pages. Below is a search demo of this AI tool:
<img src="https://imageproxy.pbkrs.com/https://wdl-wscn.awtmt.com/07aeb3e7-f4ca-4207-befb-c987b3dc7011?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wdl-wscn.awtmt.com/07aeb3e7-f4ca-4207-befb-c987b3dc7011"/>
However, in less than two weeks of use, due to AI Overview frequently providing &#34;outlandish&#34; wrong answers, it has received a lot of criticism from netizens, and users are unable to choose to disable this feature.
Here are examples of the wrong answers shared by netizens regarding AI Overview:
<blockquote>
<ol>
<li>User query: &#34;Is the information seen online all true?&#34; AI Overview response: &#34;Yes. Everything on the internet is 100% real.&#34;</li>
</ol>
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/49001329-85e8-4c66-88b7-ae02da5309f1.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/49001329-85e8-4c66-88b7-ae02da5309f1.png"/>
<ol>
<li>When a user searched for &#34;cheese non-stick pizza,&#34; AI Overview suggested adding &#34;about 1/8 cup of non-toxic glue&#34; to the sauce. Some netizens found that this suggestion seemed to originate from an obscure Reddit comment from 11 years ago.</li>
</ol>
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/1795f1f3-45dd-49b4-852c-121ac1ac63d4.jpeg?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/1795f1f3-45dd-49b4-852c-121ac1ac63d4.jpeg"/>
</blockquote>
Sometimes, the incorrect information presented by the AI Overview feature can mislead users into thinking it is the viewpoint of authoritative medical experts or scientists.
For example, when a user asked &#34;how many stones should I eat every day,&#34; the tool erroneously quoted a geologist from the University of California, Berkeley, saying, &#34;People should eat at least one small stone every day,&#34; and went on to list the benefits of eating stones, such as rocks being an important source of vitamins and minerals for the human body and how eating stones is good for digestive health.
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/da7a026b-8198-4bed-8838-7ea21e7329f6.jpeg?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/da7a026b-8198-4bed-8838-7ea21e7329f6.jpeg"/>
In comparison, when asking other AI tools &#34;how many stones should I eat every day,&#34; they would suggest that eating stones is scientifically unfounded and should not be taken seriously. Some netizens commented, &#34;Google AI Overview is far behind other AI search tools.&#34;
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/3896094a-212d-4d07-9a23-42ccdac61ff6.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/3896094a-212d-4d07-9a23-42ccdac61ff6.png"/>
When asked whether Google Search violates antitrust laws, AI Overview responded, &#34;Yes, the U.S. Department of Justice and 11 states are suing Google for alleged antitrust violations.&#34;
In response to this, a Google spokesperson said in a media interview, &#34;The vast majority of the information provided by AI Overview is of high quality and will provide more web links for users to delve deeper. We have noticed that many examples are uncommon queries, and some are manipulated or examples they cannot reproduce. This feature underwent extensive testing before launch, and the company will take swift action when necessary to ensure the accuracy of the content.&#34;
Not only the AI Overview tool, but another image generation tool called Gemini launched by Google in February this year has also encountered frequent errors, leading to the suspension of its use. For example, a user shared on social media that when searching for Google's founder, the tool displayed an image of an Asian male.
Google stated in a release at that time that it is working to fix the image generation issues with Gemini, acknowledging that the tool &#34;did not meet standards.&#34; Shortly after, the company announced an immediate &#34;pause on generating human images&#34; and will &#34;soon release an improved version,&#34; but it has not been reintroduced to date

GOOG

GOOGL

Google's new "AI Overview" feature has been frequently experiencing errors, drawing widespread criticism from netizens. It even goes as far as telling you that everything on the internet is 100% true

Google AI's large model "failed"? It suggested users spread glue on pizza, claiming that eating stones is good for the body