The Web is Dead, Long Live the Web


The original version of the Web based on high-quality organic hypertext links is long gone. It has been largely replaced with captive silos of information targeted at keeping users within walled gardens to the maximum extent with links to external content being marginalized in order to keep the user engagement the highest. Social networks are prime examples of such websites, but there are also many other vertical silos targeted at real estate, jobs, cars, general as well as programming and mathematical questions and answers, encyclopedic content in open form as well as general topical forums etc.

The vast diversity and scale of the original open Web is gone, as can be clearly seen by the (rapidly) declining quality of Google results that have been concentrated to a decreasing number of sites. Wikipedia is one of the prime examples of such sites, but there are others such as Quora, Reddit, Stack Overflow etc.

In addition, the shift to mobile has replaced open webpages containing links with captive closed apps. The mobile transition to smart phones as primary computing devices can really be viewed as a counter-revolution to the original open Web. In such a world the original value and point of search engines as information systems focused on returning results as high quality web pages where keywords appear has lost meaning and relevance. Instead, the focus has long shifted to giving users direct answers to their queries phrased as questions quickly so they can continue with their activities.

Another key problem and issue from the very beginning of the Web has been the lack of large scale search APIs. Google has been very closed in providing search APIs for their results, which are of limited scale and very expensive, as they chose (not surprisingly) to keep their crown jewels for themselves.

YottaAnswers solves the above problems by a novel smart Very Large Scale AI (VLSAI) system capable of understanding questions and giving direct answers to them, as opposed to the now legacy technology of just matching keywords. Our system is totally focused and completely dedicated to protecting user privacy by rejection of cookies and other forms of user tracking.

The scale of the system is very large, in billions of questions and answers, which is crucial for the value and utility we provide to our users. The resources required for our system are in (stark) contrast to the contemporary AI efforts that require enormous costs such as e.g. OpenAI GPT-3 transformer model with 175B parameters that requires weeks of running and millions of dollars for a single training run. Such costs are prohibitive even for the largest players such as Google, which can be clearly seen by the very small scale of Google search queries that are directly answered (as opposed to returning links based on keywords).

Our system is much more efficient, requiring resources that are several orders of magnitude smaller, clearly demonstrating that the gigantic sizes of contemporary AI models such as transformers are unnecessary and misguided.

Our business model is based on licensing our data and technology as opposed to massive ever-increasing advertising that is highly disruptive to users. In addition, YottaAnswers is technology with no sign up, no tracking, no cookies, no chat bots, no disturbing of any kind – just direct answers, free of charge.