As a former web developer and current full time online privacy and security advocate, I'm always looking for the best tools to browse the web securely. I figured the new writing contest launched by Publish0x would be a good opportunity to talk about the privacy aspect of Presearch, to see how secure it actually is (or will be), and how it compares to its competitors.
How does a search engine work?
Before we get into Presearch, I thought it could be interesting and worthwhile to start with a brief explanation (or reminder) of how a search engine works. Without going too deep into technical details, here's a simple breakdown.
First of all, a search engine typically uses three tools to look for and store information about everything that exists on the internet:
- Web crawlers: an army of bots that continuously browse the web looking for new things, and collecting some specific information on web pages as they do so.
- A search index: think of it as a huge database storing every single page found by the crawler and associating them with things like keywords, locations, etc
- Search algorithms: this gets more specific for each engine, but they will basically decide how to rank pages, based on their relevance to keywords, their popularity, content quality, date of publication, etc.
Now that we have a search engine, let's see how we can get results when you're looking for something. When you type words in a search bar, it sends a request to a server where the algorithms will be trying to find results for you:
- First, the request is parsed, meaning each word is analysed, alone and in combination with the other ones. The request always contains some information such as your IP address, location, language, the date, etc and can sometimes even carry more private things (like your search history) that will be parsed and stored somewhere for future use (I'll let you guess which engine does that).
- The algorithms will use all of this information -both your keywords and the extra data- to try to guess your intent and come up with results they assume will match your expectations. This means that if you live in California and you just type "election results" the day after a local election, a good algorithm should expect that you want to find out yesterday's local election results in California, not Poland's 2005 presidential election results.
- They will then go through the index to get all the pages they think you'll want to see and send this list of results back to the server, which will in turn display it in your browser.
Is Presearch's search engine secure?
Presearch claims to be creating a decentralized search engine, so what does that entail?
While the algorithms of almost every search engine out there are usually run on the company's servers, where they can do pretty much anything they want with all the data we're feeding them, Presearch has created a node based search engine, where the requests go through nodes hosted by the community, aka people like you and me.
Well, that's cute, but does that mean I'm just feeding my data to random people now? Sounds even worse...
Enter the Node Gateway. This is Presearch's bridge between the end user (you) and the node workers. Its main job is to strip the query from all the PII -Personally Identifiable Information- it contains. Simply put, when you type keywords in the search bar and hit "Enter", the query is then sent to the web server, as usual, and then sent directly to the Node Gateway. Here, your IP address, location, browser, device, etc will simply be removed. The new query, that now only contains what someone is looking for, and not at all any information about who is looking for it, is sent to a random available node on the network. This query will also be sent to Presearch's advertising API to retrieve relevant ads, based on the keywords.
This node will look for results and send them back to the gateway, which will in turn send everything back to the server, with the ads, for you to see. This is all broken down in the chart below, taken form Presearch's White Paper.
It's important to note that, even though Presearch advertises a decentralized search engine, it is not 100% decentralized yet. The Nodes were launched as a first step at the end of JAnuary, but the rest is still centralized and run by Presearch as company. Once the node network used for search proves to be sustainable, robust and secure, it will be time for the Node Gateway to be decentralized, and this will be another challenge because of the user information the gateway sees and handles with each query. It remains unclear how exactly this will be achieved, as the White Paper only says on the matter: "Gateway operators will eventually consist of a smaller network of trusted providers that the Presearch community may select". This "selection" will inevitably raise more questions of privacy.
In its final form, Presearch aims to be fully decentralized, with node workers running all the different parts of the system, for the whole thing to be working smoothly and be 100% handled by the community.
DuckDuckGo, Qwant and StartPage are all privacy focused search engines, and they basically all do it the same way:
- They do not store your search history,
- They do not set up trackers or cookies that could then be used for advertising purposes,
- They do not store any identifying information, that could be used to create an advertising profile based on your searches.
The main takeaway is that you can't sell/leak/lose private information if you do not store said information. These 3 are very good options for private searches, and have been dominating the market for a long time. The difference will mostly be seen on the results you get, as their algorithms are all different.
Obviously, if you are a registered Presearch user, either to receive tokens for searching or to run a node, they will store the private information you give them when you create your account: name, email address, password (encrypted), tokens, wallet address, etc, but that's always the case when you create an account somewhere.
I will definitely keep an eye on Presearch as the roadmap keeps unfolding and more and more features and layers get decentralized. If they manage to achieve 100% decentralization, they can definitely be on top of the privacy search engine market, with the added bonuses of the PRE token, the associated tokenomics system and the community dynamics. Whether you want to use Presearch or not, keep in mind that a private search engine alone does not guarantee online privacy. Using a privacy focused web browser (like TOR, Brave or Firefox) and a VPN (like the free ProtonVPN) will give you the extra layers of security needed for a safe and private browsing experience.
Any kind of support, even a simple "like", is truly appreciated and helps me grow and create more quality content for you to enjoy.
Buy Me A Coffee: https://www.buymeacoffee.com/piliz
Enjoy 1 random free stock (no deposit required) by signing up on Robinhood: https://join.robinhood.com/nathaln-81f11a
Chart credits: All rights reserved to Presearch