Computer science: by accident?
I was studying mathematics at Normale (French graduate school) at the end of the 1970s. I went to enrol my partner at the university and I saw the poster for Jean Vuillemin's DEA (post-graduate degree). I began the computer science classes without really knowing what it was, and I had to learn how to program on punched cards. I was also working with Pierre Boulez's IRCAM (French Institute for Research and Coordination in Acoustics/Music): the centre was open at night, and made some remarkable encounters possible. There I applied what I learned in Gérard Huet's classes: he came by one day and, to his surprise, discovered my work on the screen! I then did my thesis, which focused on formal proofs, under his supervision at Inria. I remember great flexibility within defined projects; we could go from one team to another, take part in very different projects. Everyone could also give each other the necessary tools in order to develop their research.
I had not really considered the world of industry: I was doing computer science because I liked it and was trying to create some beautiful things that were easy to understand. I must have spent as much time on the layout of my thesis as I did on the thesis itself! Following my PhD, I began working with Jean Vuillemin on the theory of integrated circuits, and I was very interested in the slides used to present the classes: aesthetics is essential as it builds things, it determines what is important and what is not. Another great aspect of the institute was the opportunity to travel and discover other methods and other professors. You need to organise these types of breaks, because if you do the same thing all the time your mind becomes hard-wired and no longer leaves its system. You can create new layers, but you stay in the same niche. It is by changing system that disruptive technologies appear; it is by freeing ourselves from the things we know that we innovate. That is what I did during a mission to Reunion at the time.
Innovation, from Rocquencourt to California
The programming and effective implementation aspect became essential. I did not want my work to remain theoretical. When I came back from Reunion in 1984, the Macintosh had been released. Apple had given two to the institute, who did not know what to do with them. My colleagues put one on my desk, thinking that it would keep me entertained! It was a true revolution for me: at Inria we were still working with screens and characters, without windows and without graphics. The tool was still not very powerful, but offered both new work possibilities and new problems. The fact of having to build a user interface by writing lines of code involved an extremely significant amount of extra work. The result was certainly more pleasant and could be directly used by others, but the time wasted for a researcher was enormous. I therefore had the idea of building all of the graphics part with the help of a graphics editor. That way, there was no longer any programming required to create the windows, buttons, menus, etc. All that remained was to "connect" the interface elements to the different functionalities of the program. Once again, I found the solution to do this entirely graphically by using object-oriented programming concepts. As a result I created the tool, called SOS Interface, that greatly appealed to the other researchers but also to Apple...and I told myself that it could not just remain within a research institute.
That way, there was no longer any programming required to create the windows, buttons, menus, etc. All that remained was to "connect" the interface elements to the different functionalities of the program. Once again, I found the solution to do this entirely graphically by using object-oriented programming concepts. As a result I created the tool, called SOS Interface, that greatly appealed to the other researchers but also to Apple...and I told myself that it could not just remain within a research institute.
The president at the time, Alain Bensoussan, was totally taken up with the creation of ILOG so we agreed to develop this disruptive technology on a wider scale; he left me the intellectual property of the tool and I resigned from Inria. I spent a year in my garage in order to turn it into a truly marketable product, which I then distributed in the United States under the name Interface Builder. I presented it at MacWorld Expo in San Francisco. A product marketing representative from NeXT, the company that had just been created by Steve Jobs following his departure from Apple, invited me to come and see them. And so I met Steve, who as usual was a little bit aggressive and who, after congratulating me, dismissed me! Then, in the car park, the product marketer came to look for me for a second interview. Steve, who was now very welcoming, was waiting for me on a sofa and asked me when I wanted to start working with him. I told him I was interested but that first of all I needed a two month holiday! But Steve really liked different people and attitudes and, in the United States such a request was obviously pretty surprising! So I left for Laos.
Mac OS and the digital revolution
When I returned, I agreed to implement my solution at NeXT for six months, on one condition: absolute freedom with regard to my return trips to Paris where my partner had just found a job. Steve accepted – strangely - and every other weekend I went back to Paris. In the end I stayed there 10 years. It was the most enjoyable period, even more so than at Apple afterwards: NeXT was a small company where lots of things were starting up, and when I joined there were barely 30 of us. Everyone worked day and night; Steve would turn up at 3 am with a new idea that had to be put in place immediately. We were dreamers, we wanted to change the world. There was a closeness, particularly with Steve, that was extremely productive. Apple was, and still is, a machine with many internal political issues. At NeXT, and with Bertrand Serlet in particular, we built what was to become Mac OS and, later, the iPhone operating system.
We put the entire operating system in place, really putting the emphasis on building the interface. We created the most practical area in which to create programs, a truly avant-garde space but one that required very costly machines. There was significant critical acclaim, but little commercial success. The CERN, which had subsequently implemented the Web, was using NeXT machines, but they were mostly of interest to bankers, who needed flexible interfaces in order to speed up financial operations. The work of the CERN and Tim Berners-Lee was of greater interest, and it reached its full potential at the end of the 1990s.
For my part, I left three months prior to Apple's takeover of NeXT in 1996, as the banking approach was not for me. At the time, Steve was mainly tied up with Pixar - I saw all of the creation stages of Toy Story. In some ways, his mind seemed to be elsewhere; and I needed to return to Europe. Incidentally, I had three months in which to use my stock options following my departure from NeXT. I had not taken care of it at all. It was Steve who rang me at 3 am the day before the announcement of Apple's takeover of NeXT to tell me the news, a week before the expiry date of my stock options! In cases that like, you act quickly!
The invention of the iPhone and the cloud
Bertrand Serlet was part of the gradual return of Steve Jobs to Apple, and was pivotal in the integration of the NeXT system on OS 10. Meanwhile, I had taken a particular interest in the Internet, which was just starting up, and with Louis Monier's Altavista search engine in particular. In 2000, before the iPod was released, Steve came to Paris to test other waters. He had always been fond of developing external hooks, so I told him that he had no other choice but to make a mobile phone. He did not really understand the idea: the Americans had been the first on the mobile market, but had not been making any headway for some time. They had been at a standstill due to old-fashioned telephones and networks that were too weak. Europe was producing attractive telephones, however they were not very functional and their network access was deliberately restricted. The manufacturers wanted to control this access and had blocked it - with the WAP - so that they could remain in control. In Japan, on the other hand, people were already following GPS on their telephones!
For me, the solution was to be found in the software. The Japanese did not totally master it in-house and were having problems in going further and adding more developed applications to their mobile phones. Nokia, who also had global aspirations, was stuck due to the absence of an operating system similar to that of computers. So I gave Steve this idea: a telephone with the equivalent of Mac OS 10, whose software would be totally controlled by Apple, and which we could equip with various applications. We all know how big this idea became.
Steve immediately found the name "iPhone", but was not convinced of the validity of the project, especially as he wanted to ensure he controlled the entire music market with the iPod before trying his hand at telephony. I set up a team of 25 people in Paris to create a relay system between the computer and the other future tools: I really was convinced that the tools were going to multiply, that they would need to be interconnected and ensure that the users were no longer worried about data issues. Everything - contacts, appointments, etc. - had to be available. That was the key of what was then to become the cloud, and it was done in Paris!
In 2006, Steve decided to make this telephone: the iPod was on the decline and he had the brainwave of producing a laptop without a keyboard - something that was unthinkable at the time, and which caused quite a stir at Apple. We were a secret project, barely known to Apple France. When Steve made the decision to put the project into production, he closed everything in Paris and I had to go back to the United States in order to continue. I did not want to go, especially as the really innovative part - which interested me the most - was almost complete.
A look back in pictures
I then participated in the creation of an online image encyclopaedia, Fotopedia: we managed to collect a lot of photos, but the economic model was somewhat shaky. At the same time as Apple brought out the iPad we had the idea of creating Heritage, a photographic application that could use our funds. The application, certified by UNESCO, met with great success, but did not manage the transition to a paid model. So we switched to a magazine medium in order to get users back on a regular basis. We survived - thanks to advertising - until the arrival of Facebook, which caused advertising prices to plummet. Without sufficient margins to adapt to Facebook, we closed. This is the principle behind any start-up: it evolves, changes its strategic direction, product, customer...and must accept the law of competition!
Today, my partner and I have an environmental foundation, Fondation Iris, which supports numerous projects aimed at raising awareness. For example, we went on a mission to Papua New Guinea with photographers, illustrators but also ornithologists and seabed specialists to make films and photo-montages, etc. The foundation also supports projects in organic agriculture, site protection and biodiversity.
Jean-Marie Hullot in a few dates
- 1981 : PhD in computer science at Orsay under the supervision of Gérard Huet
- 1979-1986 : researcher at IRIA then Inria
- 1986-1996 : begins at NeXT
- 1996 : co-founding RealNames, URL translation service
- 2000-2005 : CTO of Apple's Applications Department
- 2014 : founding of Fotopedia