Crazy ideas
Ideas are many. It's the execution that counts.
But time and opportunity are limited.
You can't make everything you think of.
These ideas stayed on the page.
Time is closing in on my arc of prediction.
My earlier ideas were 15-20 years ahead.
My later ideas are in the now.
Which means the world I've been
imagining for 20 years is finally, tantalisingly, almost here.
But time and opportunity are limited.
You can't make everything you think of.
These ideas stayed on the page.
Time is closing in on my arc of prediction.
My earlier ideas were 15-20 years ahead.
My later ideas are in the now.
Which means the world I've been
imagining for 20 years is finally, tantalisingly, almost here.
2006
2006
GAME
HUNGARY: 1943
A deeply personal project. I visited Hungary with my wife while we were dating. My Grandmother came from there, fleeing the Nazis to England just before they came knocking and the visas ended. I imagined creating a recreation of the city at that time, the segregation how it felt. I wanted to fill the world with autonomous agents - with goals, hopes, desires - each with a unique perspective on their own life, the lives of those around them and the circumstances they've found themselves caught up in.
I even drafted a letter to Steven Speilberg to get involved and fund the project.
To my shame I never sent it.
I was too full of fear - not perhaps of failing.. but of succeeding...
I even drafted a letter to Steven Speilberg to get involved and fund the project.
To my shame I never sent it.
I was too full of fear - not perhaps of failing.. but of succeeding...
2006
2006
GAME
Pike
Stupid name but a prescient reading of a world gone social, capitalism and populism at their worst.
Before smartphones, before Facebook, influencers and Twitter, I imagined a world of listen and respond product creation that turned cultural appropriation into high-concept liberal leaning artform. Oh and it meant beating up Nazis.
I'm not sure that I ever really fleshed out the actual gameplay, for this one!
READ THE DESIGN DOC
Before smartphones, before Facebook, influencers and Twitter, I imagined a world of listen and respond product creation that turned cultural appropriation into high-concept liberal leaning artform. Oh and it meant beating up Nazis.
I'm not sure that I ever really fleshed out the actual gameplay, for this one!
READ THE DESIGN DOC
2004
2004
GAME
10 Minutes
THIS IS THE BIG ONE
This game idea has never left me. I obsessed over it. I would lay awake at night imagining playing it. I invented all the custom tools you'd need to create it. My fingers twitched through playing it from start to finish countless times.
The idea has never left me. I've wrapped it in so many different stories, tweaked the gameplay from strategic, to arcade, to casual and arthouse emotional.
While the idea has never been made, I still feel that the mechanics solve a problem that has plagued games for years.
1: Games for grown-ups. Not mature meaning swearing, nudity and blood, but mature meaning considered, cogent, intelligent.
2: Narrative as gameplay: Not boring exposition
3: Fun for fingers: but not a skill test
Back in 2003, before the Wii, before iPhones and before Indie games, those three principles were utterly incomprehensible. I persisted in selling it in. I failed. I'd still like to make this game. It builds on elements of ideas past: on-rails action synced with music, an ultra-dense shifting crowd of people that create pathways and AI to generate the fun in realtime. The new element that ties it all together: control over time.
Here's the first, SERIOUS version of the idea
This take is much more casual 5 years later for Wii (I didn't give up on you, idea!)
This game idea has never left me. I obsessed over it. I would lay awake at night imagining playing it. I invented all the custom tools you'd need to create it. My fingers twitched through playing it from start to finish countless times.
The idea has never left me. I've wrapped it in so many different stories, tweaked the gameplay from strategic, to arcade, to casual and arthouse emotional.
While the idea has never been made, I still feel that the mechanics solve a problem that has plagued games for years.
1: Games for grown-ups. Not mature meaning swearing, nudity and blood, but mature meaning considered, cogent, intelligent.
2: Narrative as gameplay: Not boring exposition
3: Fun for fingers: but not a skill test
Back in 2003, before the Wii, before iPhones and before Indie games, those three principles were utterly incomprehensible. I persisted in selling it in. I failed. I'd still like to make this game. It builds on elements of ideas past: on-rails action synced with music, an ultra-dense shifting crowd of people that create pathways and AI to generate the fun in realtime. The new element that ties it all together: control over time.
Here's the first, SERIOUS version of the idea
This take is much more casual 5 years later for Wii (I didn't give up on you, idea!)
2003
2003
GAME
Omerta
I have always loved breaking genres and pushing game mechanics to solve things that didn't seem solvable. Point and click adventure games always seemed lonely. I wished there could be a point and click that was social, competitive and team based.
The outcome was Omerta. A series of fetch quests and resource gathering as a team sport.
The idea was to gain control of the town, by controlling all the key buildings. You do this by taking possession of Briefcases, each one holds scandalous documents that give you leverage over one building's owner, putting them in your pocket. Throw in gunplay, chains of items to grab hold of in sequence and fetch quests while the bullets are flying everywhere in a prohibition-era setting. Still a fun idea, but still riddled with holes too!
READ THE DESIGN DOC
The outcome was Omerta. A series of fetch quests and resource gathering as a team sport.
The idea was to gain control of the town, by controlling all the key buildings. You do this by taking possession of Briefcases, each one holds scandalous documents that give you leverage over one building's owner, putting them in your pocket. Throw in gunplay, chains of items to grab hold of in sequence and fetch quests while the bullets are flying everywhere in a prohibition-era setting. Still a fun idea, but still riddled with holes too!
READ THE DESIGN DOC
2002
2002
GAME
The Academy
Another foray into AI for games.
I loved 4x strategy games.
I was terrible at them, but I loved the design principles.
i was also hugely into autonomous agents and object oriented programming - the ins and outs of inheritance and parent/child relationships in coding.
I imagined a game where you could directly control units, and use them as exemplars to train neural nets. The idea being, that as a drill sergeant you can train squads to behave in certain ways, and then as a commander you select the different types of troops you've created and send them out into battle to beat the enemy. By combining the different behaviours you'd created in novel ways, you craft new capabilities on the battlefield.
4X-AI fun.
I loved 4x strategy games.
I was terrible at them, but I loved the design principles.
i was also hugely into autonomous agents and object oriented programming - the ins and outs of inheritance and parent/child relationships in coding.
I imagined a game where you could directly control units, and use them as exemplars to train neural nets. The idea being, that as a drill sergeant you can train squads to behave in certain ways, and then as a commander you select the different types of troops you've created and send them out into battle to beat the enemy. By combining the different behaviours you'd created in novel ways, you craft new capabilities on the battlefield.
4X-AI fun.
2001
2001
GAME
Kung Fu
I was massively into import Kung Fu movies. The more unbelievable the action, the better. I loved the early 90's Jet Li films, like Fong Sai Yuk, stuffed full of whimsical fights that defy physics. Sequences that are so intensely creative the scope of their imagination leaves you reeling.
I wanted to make a balletic fighting game inspired by those movies. I detailed a whole Neural Network system that composed the camera and all the enemies together with the player as a graph. The idea was to let you focus your fight one individual enemy, but while you were engaging, the AI would generate flurries of blows, swipes, arrow shots from the hoard of surrounding, close combat enemies. Some few would strike you, and perhaps unbalance you imperceptibly, but many of them would just miss, striking empty space that you were in just moments before, making the whole scene look epic.
The fights would be point to point races to reach an objective, with fantastical settings, like bounding over the heads of a crowd, on the backs of water buffalo crossing a river, across bird-filled rooftops, dodging bamboo scaffolding that is constantly toppling like dominos beneath you. All the while the inverse kinematics and realtime animation generation created by the AI and framed by the AI camera, would make sure that you saw the most insane fighting, dodging and acrobatics, leaving you pumped with adrenaline and in awe at the tricks you pulled off.
I wanted to make a balletic fighting game inspired by those movies. I detailed a whole Neural Network system that composed the camera and all the enemies together with the player as a graph. The idea was to let you focus your fight one individual enemy, but while you were engaging, the AI would generate flurries of blows, swipes, arrow shots from the hoard of surrounding, close combat enemies. Some few would strike you, and perhaps unbalance you imperceptibly, but many of them would just miss, striking empty space that you were in just moments before, making the whole scene look epic.
The fights would be point to point races to reach an objective, with fantastical settings, like bounding over the heads of a crowd, on the backs of water buffalo crossing a river, across bird-filled rooftops, dodging bamboo scaffolding that is constantly toppling like dominos beneath you. All the while the inverse kinematics and realtime animation generation created by the AI and framed by the AI camera, would make sure that you saw the most insane fighting, dodging and acrobatics, leaving you pumped with adrenaline and in awe at the tricks you pulled off.
2000
2000
GAME
Unstoppable
Music ruling the gameplay now.
With a colleague, Jon Skuse, we imagined an insanely over-the-top on rails shooter, but instead of a top-down bullet hell shmup, this was a behind the shoulder 'infinite runner' style game, where all the action was tied to the music.
I invented a level creation system that would lay out the music like a sequencer in 3D space, and match reusable modules to it, like ledges, drops, platforms and enemies, so that everything synced up.
When Rez came out on the Dreamcast a year later, it felt like a dream come true. All be it in a very abstract way.
With a colleague, Jon Skuse, we imagined an insanely over-the-top on rails shooter, but instead of a top-down bullet hell shmup, this was a behind the shoulder 'infinite runner' style game, where all the action was tied to the music.
I invented a level creation system that would lay out the music like a sequencer in 3D space, and match reusable modules to it, like ledges, drops, platforms and enemies, so that everything synced up.
When Rez came out on the Dreamcast a year later, it felt like a dream come true. All be it in a very abstract way.
1999
1999
GAME
Rush
This came from a mixture of a dream and an experience. I was in Edinburgh for the New Year celebrations around the castle. In the crowd I met a beautiful American girl, at midnight, with the fireworks bursting around us, we ran, hand in hand through the dense crowd.
Around the same time I woke up from a nightmare. I'm trapped in a prison cell, suddenly all the cells burst open and we flee, scattering in different directions, trying to escape.
From this came the big idea for Rush, a kind of story-driven endless runner (before they had ever been invented) where instead of a fixed path, or a road, the way ahead opened - dynamically in ever changing milling of a dense crowd of people.
This idea of blending emergent crowd dynamics, with AI based pathway generation to give you the most pleasurable experience formed part of a lexicon of game mechanics that I applied to pitches over and over. No one ever bought it, but I still love it.
Around the same time I woke up from a nightmare. I'm trapped in a prison cell, suddenly all the cells burst open and we flee, scattering in different directions, trying to escape.
From this came the big idea for Rush, a kind of story-driven endless runner (before they had ever been invented) where instead of a fixed path, or a road, the way ahead opened - dynamically in ever changing milling of a dense crowd of people.
This idea of blending emergent crowd dynamics, with AI based pathway generation to give you the most pleasurable experience formed part of a lexicon of game mechanics that I applied to pitches over and over. No one ever bought it, but I still love it.
1997
1997
DEVICE
Sprayless
Roads are pretty dangerous, especially when they are wet. Driving on the motorway (freeway to you Americans) at 70mph in a haze of tyre spray seemed to me as a new driver to be a hazard worth solving.
So I remembered that clever little trick of rubbing a balloon and holding it near a tap could divert the water flow, and bend it.
I know, I stupidly thought. What about using static electricity to catch all the spray coming off the tyres. I imagined getting the patent and getting governments the world over to insist on them being fitted as a safety feature. An engineering friend of mine calculated that the the power needed to divert the volume of water generated per second was so big, it would probably blow the petrol tank, that I wound up the company.
I'm probably still listed as an ex-director of Sprayless Ltd.
So I remembered that clever little trick of rubbing a balloon and holding it near a tap could divert the water flow, and bend it.
I know, I stupidly thought. What about using static electricity to catch all the spray coming off the tyres. I imagined getting the patent and getting governments the world over to insist on them being fitted as a safety feature. An engineering friend of mine calculated that the the power needed to divert the volume of water generated per second was so big, it would probably blow the petrol tank, that I wound up the company.
I'm probably still listed as an ex-director of Sprayless Ltd.
1996
1996
ARTIFICIAL INTELLIGENCE
SOULS & SELF-LEARNING NETWORKS
I'm still pretty annoyed about this one.
I obsessed over hand-drawn network diagrams, which I took to my professor out-of-hours.
The big idea? Get one neural-net to train a second one, and create a feedback loop where they continually make each other better.
My professor took one look at it and his eyes glazed over. Before I could even get into its stochastic nature I was ushered out. I think he gave me 5 minutes. I admit that I let this beat me. I never stopped thinking about my two networks outcompeting each other, because I fell in love with the idea.
But at the time, there were no terabyte repositories of training data and compute power was very limited. I was stuck writing a crossword solver for my dissertation, and I abandoned the concept, except to noodle on it with various colleagues over the years.
Inspiration for it came because I was reading a lot about quantum effects that affect the way our brains process things, given the scale and proximity of neurones. I imagined that we had additional processing, occurring in N dimensional space, like a much bigger brain that sits through/beyond/outside/within our own brain. It made it possible in my mind to reconcile some weird precognitive dreams I'd been having, if part of the brain lived in a non-euclidean space.
So I took that as my starting point and imagined a second larger neural net that was the "true" network, spinning out generations of smaller task solving networks. Essentially the larger net would evaluate the progress of the smaller net and both together would be run through many training cycles.
I obsessed over hand-drawn network diagrams, which I took to my professor out-of-hours.
The big idea? Get one neural-net to train a second one, and create a feedback loop where they continually make each other better.
My professor took one look at it and his eyes glazed over. Before I could even get into its stochastic nature I was ushered out. I think he gave me 5 minutes. I admit that I let this beat me. I never stopped thinking about my two networks outcompeting each other, because I fell in love with the idea.
But at the time, there were no terabyte repositories of training data and compute power was very limited. I was stuck writing a crossword solver for my dissertation, and I abandoned the concept, except to noodle on it with various colleagues over the years.
Inspiration for it came because I was reading a lot about quantum effects that affect the way our brains process things, given the scale and proximity of neurones. I imagined that we had additional processing, occurring in N dimensional space, like a much bigger brain that sits through/beyond/outside/within our own brain. It made it possible in my mind to reconcile some weird precognitive dreams I'd been having, if part of the brain lived in a non-euclidean space.
So I took that as my starting point and imagined a second larger neural net that was the "true" network, spinning out generations of smaller task solving networks. Essentially the larger net would evaluate the progress of the smaller net and both together would be run through many training cycles.
1993
1993
GAME
You and Me
A co-operative two player, side-scrolling platformer, starring Billy and Betty, wanna be circus stars.
They had all kinds of acrobatic moves that they could perform together, so it was part rhythm action and part classic '90s jumping and avoiding enemies and spike pits.
It was imagined as 2.5D - with fully 3D backgrounds, and a path that wove in and out of view, but it was all played on a single 2D plane, side on, like Mario and Sonic at the time.
They had all kinds of acrobatic moves that they could perform together, so it was part rhythm action and part classic '90s jumping and avoiding enemies and spike pits.
It was imagined as 2.5D - with fully 3D backgrounds, and a path that wove in and out of view, but it was all played on a single 2D plane, side on, like Mario and Sonic at the time.
1993
1993
DEVICE
Next Please!
The Amiga could do all kinds of great video stuff.
I imagined a system that could be used in doctors' waiting rooms to show you whose turn it was next, and where they had to go.
Sounds stupid, beacuse they are everywhere now.
But back in the 90's there would be a buzzer and the receptionist would call out "Mr. Smith to room 3 please".
The business model was to provide them for free, but earn advertising revenue from each one installed.
I imagined a system that could be used in doctors' waiting rooms to show you whose turn it was next, and where they had to go.
Sounds stupid, beacuse they are everywhere now.
But back in the 90's there would be a buzzer and the receptionist would call out "Mr. Smith to room 3 please".
The business model was to provide them for free, but earn advertising revenue from each one installed.
1992
1992
GAME IDEA
Deep Trouble
An underwater 3D adventure game, where you can tag various creatures with a chip that hacks into their brain and so you can pilot them first person.
Essentially a mission-based submarine game, where you play at varying scales from a tiny fish, all the way up to a giant whale. Each size you play at gives you different gameplay. Larger fish can dive to deeper depths, while getting bigger means you no longer fit into smaller spaces.
Essentially a mission-based submarine game, where you play at varying scales from a tiny fish, all the way up to a giant whale. Each size you play at gives you different gameplay. Larger fish can dive to deeper depths, while getting bigger means you no longer fit into smaller spaces.
1991
1991
GAME IDEA
Fury
A sprawling point and click adventure, that blended run+gun mechanics with puzzle solving.
This was the big one, the first idea I fell in love with.
I would obsess over this idea for over 10 years in one form or another.
I also got to pitch it to Psygnosis in Liverpool, in person at the age of 14, and later on in a revised form to Ocean.
This was the big one, the first idea I fell in love with.
I would obsess over this idea for over 10 years in one form or another.
I also got to pitch it to Psygnosis in Liverpool, in person at the age of 14, and later on in a revised form to Ocean.
1990
1990
GAME IDEA
Wall Builder
Where is all started.
My first ever original game idea.
You, and some fuzzy creatures are trapped in a building basement that's being slowly flooded.
A helicopter flies back and forth across the top of the screen, dropping differently shaped blocks.
Catch the blocks and fit them together to build a wall.
Every so often a wave passes across the screen, and any blocks that don't fit in properly are swept away.
Build high enough to reach the door out of the basement, setting you and the fuzzy creatures free.
Levels require you to build ever higher and to firm up your block pieces more frequently as the waves start to visit faster. I imagined it as a fast-paced arcade puzzle game, a Bubble Bobble interface for Tetris.
My first ever original game idea.
You, and some fuzzy creatures are trapped in a building basement that's being slowly flooded.
A helicopter flies back and forth across the top of the screen, dropping differently shaped blocks.
Catch the blocks and fit them together to build a wall.
Every so often a wave passes across the screen, and any blocks that don't fit in properly are swept away.
Build high enough to reach the door out of the basement, setting you and the fuzzy creatures free.
Levels require you to build ever higher and to firm up your block pieces more frequently as the waves start to visit faster. I imagined it as a fast-paced arcade puzzle game, a Bubble Bobble interface for Tetris.
2003
2003
GAME
Omerta
I have always loved breaking genres and pushing game mechanics to solve things that didn't seem solvable. Point and click adventure games always seemed lonely. I wished there could be a point and click that was social, competitive and team based.
The outcome was Omerta. A series of fetch quests and resource gathering as a team sport.
The idea was to gain control of the town, by controlling all the key buildings. You do this by taking possession of Briefcases, each one holds scandalous documents that give you leverage over one building's owner, putting them in your pocket. Throw in gunplay, chains of items to grab hold of in sequence and fetch quests while the bullets are flying everywhere in a prohibition-era setting. Still a fun idea, but still riddled with holes too!
READ THE DESIGN DOC
The outcome was Omerta. A series of fetch quests and resource gathering as a team sport.
The idea was to gain control of the town, by controlling all the key buildings. You do this by taking possession of Briefcases, each one holds scandalous documents that give you leverage over one building's owner, putting them in your pocket. Throw in gunplay, chains of items to grab hold of in sequence and fetch quests while the bullets are flying everywhere in a prohibition-era setting. Still a fun idea, but still riddled with holes too!
READ THE DESIGN DOC
2002
2002
GAME
Another foray into AI for games.
I loved 4x strategy games.
I was terrible at them, but I loved the design principles.
i was also hugely into autonomous agents and object oriented programming - the ins and outs of inheritance and parent/child relationships in coding.
I imagined a game where you could directly control units, and use them as exemplars to train neural nets. The idea being, that as a drill sergeant you can train squads to behave in certain ways, and then as a commander you select the different types of troops you've created and send them out into battle to beat the enemy. By combining the different behaviours you'd created in novel ways, you craft new capabilities on the battlefield.
4X-AI fun.
I loved 4x strategy games.
I was terrible at them, but I loved the design principles.
i was also hugely into autonomous agents and object oriented programming - the ins and outs of inheritance and parent/child relationships in coding.
I imagined a game where you could directly control units, and use them as exemplars to train neural nets. The idea being, that as a drill sergeant you can train squads to behave in certain ways, and then as a commander you select the different types of troops you've created and send them out into battle to beat the enemy. By combining the different behaviours you'd created in novel ways, you craft new capabilities on the battlefield.
4X-AI fun.
2001
2001
GAME
I was massively into import Kung Fu movies. The more unbelievable the action, the better. I loved the early 90's Jet Li films, like Fong Sai Yuk, stuffed full of whimsical fights that defy physics. Sequences that are so intensely creative the scope of their imagination leaves you reeling.
I wanted to make a balletic fighting game inspired by those movies. I detailed a whole Neural Network system that composed the camera and all the enemies together. The idea was to let you focus your fight one individual enemy, but while you were engaging, the AI would generate flurries of blows, swipes, arrow shots from the hoard of surrounding, close combat enemies. Some few would strike you, and perhaps unbalance you imperceptibly, but many of them would just miss, striking empty space that you were in just moments before, making the whole scene look epic.
The fights would be point to point races to reach an objective, with fantastical settings, like bounding over the heads of a crowd, on the backs of water buffalo crossing a river, across bird-filled rooftops, dodging bamboo scaffolding that is constantly toppling like dominos beneath you. All the while the inverse kinematics and realtime animation generation created by the AI and framed by the AI camera, would make sure that you saw the most insane fighting, dodging and acrobatics, leaving you pumped with adrenaline and in awe at the tricks you pulled off.
I wanted to make a balletic fighting game inspired by those movies. I detailed a whole Neural Network system that composed the camera and all the enemies together. The idea was to let you focus your fight one individual enemy, but while you were engaging, the AI would generate flurries of blows, swipes, arrow shots from the hoard of surrounding, close combat enemies. Some few would strike you, and perhaps unbalance you imperceptibly, but many of them would just miss, striking empty space that you were in just moments before, making the whole scene look epic.
The fights would be point to point races to reach an objective, with fantastical settings, like bounding over the heads of a crowd, on the backs of water buffalo crossing a river, across bird-filled rooftops, dodging bamboo scaffolding that is constantly toppling like dominos beneath you. All the while the inverse kinematics and realtime animation generation created by the AI and framed by the AI camera, would make sure that you saw the most insane fighting, dodging and acrobatics, leaving you pumped with adrenaline and in awe at the tricks you pulled off.
2000
2000
GAME
Music ruling the gameplay now.
With a colleague, Jon Skuse, we imagined an insanely over-the-top on rails shooter, but where all the action was tied to the music.
I invented a level creation system that would lay out the music like a sequencer in 3D space, and match reusable modules to it, like ledges, drops, platforms and enemies, so that everything synced up.
When Rez came out on the Dreamcast a year later, it felt like a dream come true. All be it in a very abstract way.
With a colleague, Jon Skuse, we imagined an insanely over-the-top on rails shooter, but where all the action was tied to the music.
I invented a level creation system that would lay out the music like a sequencer in 3D space, and match reusable modules to it, like ledges, drops, platforms and enemies, so that everything synced up.
When Rez came out on the Dreamcast a year later, it felt like a dream come true. All be it in a very abstract way.
1999
1999
GAME
This came from a mixture of a dream and an experience. I was in Edinburgh for the New Year celebrations around the castle. In the crowd I met a beautiful American girl, at midnight, with the fireworks bursting around us, we ran, hand in hand through the dense crowd.
A nightmare, I'm trapped in a prison cell, suddenly all the cells burst open and we flee, scattering in different directions, trying to escape.
From this came the big idea for Rush, a kind of story driven endless runner (before they had ever been invented) where instead of a fixed path, or a road, the way ahead opened - dynamically in ever changing milling of a dense crowd of people.
This idea of blending emergent crowd dynamics, with AI based pathway generation to give you the most pleasurable experience formed part of a lexicon of game mechanics that I applied to pitches over and over. No one ever bought it, but I still love it.
A nightmare, I'm trapped in a prison cell, suddenly all the cells burst open and we flee, scattering in different directions, trying to escape.
From this came the big idea for Rush, a kind of story driven endless runner (before they had ever been invented) where instead of a fixed path, or a road, the way ahead opened - dynamically in ever changing milling of a dense crowd of people.
This idea of blending emergent crowd dynamics, with AI based pathway generation to give you the most pleasurable experience formed part of a lexicon of game mechanics that I applied to pitches over and over. No one ever bought it, but I still love it.
1997
1997
DEVICE
Roads are pretty dangerous, especially when they are wet. Driving on the motorway (freeway to you Americans) at 70mph in a haze of tyre spray seemed to me as a new driver to be a hazard worth solving.
So I remembered that clever little trick of rubbing a balloon and holding it near a tap could divert the water flow, and bend it.
I know, I stupidly thought. What about using static electricity to catch all the spray coming off the tyres. I imagined getting the patent and getting governments the world over to insist on them being fitted as a safety feature. An engineering friend of mine calculated that the the power needed to divert the volume of water generated per second was so big, it would probably blow the petrol tank, that I wound up the company.
I'm probably still listed as an ex-director of Sprayless Ltd.
So I remembered that clever little trick of rubbing a balloon and holding it near a tap could divert the water flow, and bend it.
I know, I stupidly thought. What about using static electricity to catch all the spray coming off the tyres. I imagined getting the patent and getting governments the world over to insist on them being fitted as a safety feature. An engineering friend of mine calculated that the the power needed to divert the volume of water generated per second was so big, it would probably blow the petrol tank, that I wound up the company.
I'm probably still listed as an ex-director of Sprayless Ltd.
1996
1996
ARTIFICIAL INTELLIGENCE
I'm still pretty annoyed about this one.
I obsessed over hand-drawn network diagrams, which I took to my professor out-of-hours.
The big idea? Get one neural-net to train a second one, and create a feedback loop where they continually make each other better.
My professor took one look at it and his eyes glazed over. Before I could even get into its stochastic nature I was ushered out. I think he gave me 5 minutes. I admit that I let this beat me. I never stopped thinking about my two networks outcompeting each other, because I fell in love with the idea.
But at the time, there were no terabyte repositories of training data and compute power was very limited. I was stuck writing a crossword solver for my dissertation, and I abandoned the concept, except to noodle on it with various colleagues over the years.
Inspiration for it came because I was reading a lot about quantum effects that affect the way our brains process things, given the scale and proximity of neurones. I imagined that we had additional processing, occurring in N dimensional space, like a much bigger brain that sits through/beyond/outside/within our own brain. It made it possible in my mind to reconcile some weird precognitive dreams I'd been having, if part of the brain lived in a non-euclidean space.
So I took that as my starting point and imagined a second larger neural net that was the "true" network, spinning out generations of smaller task solving networks. Essentially the larger net would evaluate the progress of the smaller net and both together would be run through many training cycles.
I obsessed over hand-drawn network diagrams, which I took to my professor out-of-hours.
The big idea? Get one neural-net to train a second one, and create a feedback loop where they continually make each other better.
My professor took one look at it and his eyes glazed over. Before I could even get into its stochastic nature I was ushered out. I think he gave me 5 minutes. I admit that I let this beat me. I never stopped thinking about my two networks outcompeting each other, because I fell in love with the idea.
But at the time, there were no terabyte repositories of training data and compute power was very limited. I was stuck writing a crossword solver for my dissertation, and I abandoned the concept, except to noodle on it with various colleagues over the years.
Inspiration for it came because I was reading a lot about quantum effects that affect the way our brains process things, given the scale and proximity of neurones. I imagined that we had additional processing, occurring in N dimensional space, like a much bigger brain that sits through/beyond/outside/within our own brain. It made it possible in my mind to reconcile some weird precognitive dreams I'd been having, if part of the brain lived in a non-euclidean space.
So I took that as my starting point and imagined a second larger neural net that was the "true" network, spinning out generations of smaller task solving networks. Essentially the larger net would evaluate the progress of the smaller net and both together would be run through many training cycles.
1993
1993
GAME
A co-operative two player, side-scrolling platformer, starring Billy and Betty, wanna be circus stars.
They had all kinds of acrobatic moves that they could perform together, so it was part rhythm action and part classic '90s jumping and avoiding enemies and spike pits.
It was imagined as 2.5D - with fully 3D backgrounds, and a path that wove in and out of view, but it was all played on a single 2D plane, side on, like Mario and Sonic at the time.
They had all kinds of acrobatic moves that they could perform together, so it was part rhythm action and part classic '90s jumping and avoiding enemies and spike pits.
It was imagined as 2.5D - with fully 3D backgrounds, and a path that wove in and out of view, but it was all played on a single 2D plane, side on, like Mario and Sonic at the time.
1993
1993
DEVICE
The Amiga could do all kinds of great video stuff.
I imagined a system that could be used in doctors' waiting rooms to show you whose turn it was next, and where they had to go.
Sounds stupid, beacuse they are everywhere now.
But back in the 90's there would be a buzzer and the receptionist would call out "Mr. Smith to room 3 please".
The business model was to provide them for free, but earn advertising revenue from each one installed.
I imagined a system that could be used in doctors' waiting rooms to show you whose turn it was next, and where they had to go.
Sounds stupid, beacuse they are everywhere now.
But back in the 90's there would be a buzzer and the receptionist would call out "Mr. Smith to room 3 please".
The business model was to provide them for free, but earn advertising revenue from each one installed.
1992
1992
GAME IDEA
An underwater 3D adventure game, where you can tag various creatures with a chip that hacks into their brain and so you can pilot them first person.
Essentially a mission-based submarine game, where you play at varying scales from a tiny fish, all the way up to a giant whale. Each size you play at gives you different gameplay. Larger fish can dive to deeper depths, while getting bigger means you no longer fit into smaller spaces.
Essentially a mission-based submarine game, where you play at varying scales from a tiny fish, all the way up to a giant whale. Each size you play at gives you different gameplay. Larger fish can dive to deeper depths, while getting bigger means you no longer fit into smaller spaces.
1991
1991
GAME IDEA
A sprawling point and click adventure, that blended run+gun mechanics with puzzle solving.
This was the big one, the first idea I fell in love with.
I would obsess over this idea for over 10 years in one form or another.
I also got to pitch it to Psygnosis in Liverpool, in person at the age of 14, and later on in a revised form to Ocean.
This was the big one, the first idea I fell in love with.
I would obsess over this idea for over 10 years in one form or another.
I also got to pitch it to Psygnosis in Liverpool, in person at the age of 14, and later on in a revised form to Ocean.
1990
1990
GAME IDEA
Where is all started.
My first ever original game idea.
You, and some fuzzy creatures are trapped in a building basement that's being slowly flooded.
A helicopter flies back and forth across the top of the screen, dropping differently shaped blocks.
Catch the blocks and fit them together to build a wall.
Every so often a wave passes across the screen, and any blocks that don't fit in properly are swept away.
Build high enough to reach the door out of the basement, setting you and the fuzzy creatures free.
Levels require you to build ever higher and to firm up your block pieces more frequently as the waves start to visit faster. I imagined it as a fast-paced arcade puzzle game, a Bubble Bobble interface for Tetris.
My first ever original game idea.
You, and some fuzzy creatures are trapped in a building basement that's being slowly flooded.
A helicopter flies back and forth across the top of the screen, dropping differently shaped blocks.
Catch the blocks and fit them together to build a wall.
Every so often a wave passes across the screen, and any blocks that don't fit in properly are swept away.
Build high enough to reach the door out of the basement, setting you and the fuzzy creatures free.
Levels require you to build ever higher and to firm up your block pieces more frequently as the waves start to visit faster. I imagined it as a fast-paced arcade puzzle game, a Bubble Bobble interface for Tetris.