'A PC is a ridiculous device." So Oracle chief executive Larry Ellison raged 20 years ago in September 1995, in a warm-up speech at an analyst conference in Paris ahead of Microsoft chief executive Bill Gates.

Gates had just launched Windows 95. It quickly became clear that the dominant computing device would be the PC running Microsoft software. Ellison also realised that Microsoft had begun to explore a move into the database sector, and hence was poised to become a deeply- pocketed threat to Oracle’s core business.

In his speech prior to Gates taking the stage, Ellison prophesied the “post-PC era”. There would be a new device, the “network computer”. It would be a low- cost ($500), simple-to-use (unlike the PC) and bare-bones machine that would serve just one purpose: to connect you to the internet. All that complex software – word processing, spreadsheets, email, web browsing, games – that used to run on PCs would instead now run on high-performance computer servers, managed by Oracle. Consumers would simply plug into these big machines across the internet, from their elegant and simple network computers.

Among others – such as Marc Andreessen of Netscape and Louis Gerstner of IBM – Eric Schmidt, chief technology officer (CTO) at Sun Microsystems, was intrigued by the concept. Under Schmidt, Sun began developing a stripped-down system to run on the new network computer. He observed: "The implications are serious: if this takes off, it will have enormous impact."

Move the clock forward 20 years to today. Ellison is now combined executive chairman and CTO of Oracle. Gates has retired but is a part-time technical advisor to Satya Nadella, the Microsoft chief executive. Schmidt is executive chairman of Alphabet, the parent company of Google. And the "post-PC era" is definitely upon us.

The "network computer" ultimately failed when, for a combination of reasons, Oracle abandoned it in 1999. However, Ellison's vision of cloud computing – running applications on huge "farms" of computer servers across the internet, rather than on desk PCs – is now dominant. His complementary vision of a simple internet connection device, a lightweight, low-cost cloud computing client, is still plausible if not particularly popular. The Google Chromebook (available in various forms from Acer, HP, Samsung and others) is the leading example.

Rather than simple devices dominating the market, we instead have highly sophisticated mobile devices such as smartphones and tablets, which combine internet connection, music player, high- definition camera, high-quality video screen and phone.

However, there is a further major transition occurring. Apparent in Asia, and not yet really obvious to us in the West, is the "post-website era". Rather than accessing information via websites, increasingly web content is available only via specific apps. Consequently, some companies are no longer bothering to maintain their websites and are instead focusing on the pertinence of their apps. Awkward web browsing is being replaced by intuitive apps.

Google is threatened by this disruption. If websites are no longer maintained, or do not even exist, then web content is not current or is even unavailable. Consequently the whole foundation of Google’s search proposition is in dire risk: Google indexes web content. Recognising the danger, in 2012 Google started encouraging app developers worldwide to consider enabling their apps so that the content inside each app could be indexed by search engines such as Google. This “deep linking” approach requires the explicit co-operation of an app developer, not unlike the explicit co-operation of webpage writers with search engines. Just last week, Google announced that nine popular apps had fully enabled their app content to become visible to Google. As a result, Google’s search results will show relevant content available – and perhaps available only – within those apps, alongside other content published on websites.

But if a Google search result shows you something which in fact is buried inside an app, does that then mean you will have to download and install that app to retrieve the content? Not necessarily: Google also announced that it will run these particular apps on its cloud servers on your behalf.

This is, potentially, a very fundamental change for the internet. No longer would you need to download, install and update apps on your smartphone or tablet. Instead, Google will run them "in the cloud" but nevertheless deliver a similar experience (look and feel) on your smart device as if you had downloaded and installed those apps. Just like Spotify, Pandora and Apple (and others) offer streaming music over the internet to your smart device, Google envisages that in fact all app content could be streamed to your device. And if that becomes the case, then your smartphone or tablet will no longer need the same computing power, battery capacity and complexity that it requires today.

The IT industry, like any other, depends on industry vision, but also capacity to put that vision into practice. The network computer was an insightful vision for the future of the internet, but Ellison’s ability to execute that vision in 1995 was unsuccessful. Twenty years later, Ellison’s vision for cloud computing is in fact with us. We may now be about to experience the other part of Ellison’s vision, in a further wave of major disruption and innovation in the internet.