1999 East Leinster under 13s triple jump bronze medalist.
Dublin
Joined June 2011
- Tweets1.6K
- Following562
- Followers187
- Likes5K
Chris Heather retweeted
HOLY MOLY: Anthropic just published what may be one of the wildest examples yet of AI-accelerated science.
Physicist Matthew Schwartz says his Claude + BootLoops setup produced 36 MANUSCRIPTS across 18 fields with 19 coauthors in just THREE MONTHS, after exploring ~400 candidate problems.
>solved an ecology equation nobody had been able to scale for 20 years
>analyzed 5.7 BILLION pairs of mutations and found evidence for gene conversion
>ported 4,452 economics papers, ~30,000 routines, and checked essentially every validatable number
>built a word-stress database covering 6,072 languages from 160,000 phonology works
And the weirdest part: Schwartz says Claude was often technically correct but scientifically uninteresting until domain experts stepped in and told it what questions actually mattered.
In physics, an “impedance mismatch” occurs when two systems each work well but are poorly matched.
In this Science Blog guest post, Harvard physicist Matthew Schwartz argues that something similar is happening with AI and science. LLMs are capable at many things, but working with them as you would with a human collaborator isn’t currently the best way to elicit their scientific strengths.
To address this mismatch, Schwartz created a toolkit for exact calculations in quantitative science. Because similar calculations often emerge in very different areas of science, Claude found connections to ecology, population genetics, and a dozen other fields, and Schwartz worked with domain experts to steer it towards interesting questions.
Read more about these projects here: anthropic.com/research/claud…
Chris Heather retweeted
Using the evidence from Auckland's upzoning, we infer that a 1% increase in housing stock reduces rents by 3.6%.
Rents fell by 21.6% and the housing stock increased by 6.9%, so the elasticity of rents to housing stock is ln(1-0.216)/ln(1+0.069) = -3.6.
I fear we can't do big ambitious things like this in Ireland anymore. So we're doomed to miss the housing targets until the system breaks. Cool idea though
A place to put the houses
You are likely familiar with our national housing targets. The latest is 50,500 homes per year.
It’s hard to put this number in context. What do 50,500 homes look like? Is it too many? Too few?
One answer is that 50,500 homes look like Cork City. We need to build one Cork City this year, and next year, and the following year, for a couple of decades. Another answer is that it looks like building a new Dublin City every four years.
The construction of 50,500 homes per year isn’t something that can be achieved by building here and there, on odd plots of land. Building that many homes demands massive ambition, planning and forward investment.
Take City Edge, for example. City Edge is (by Irish standards) an ambitious plan to turn 300 hectares of industrial land in West Dublin into high density masterplanned housing. The City Edge plan is expected to take 45 years. By 2070, it’ll have contributed 40,000 homes in total. That’s about 900 homes per year, or 1.7 per cent of our annual target. Or take Cherrywood. Cherrywood is the biggest Strategic Development Zone in the history of the state. It’s ten years in the making. When completed in the late 2020s, it’ll have yielded… 10,500 homes.
All of which is to say, we need a bigger plan for where we’re going to put all these houses.
The two paths
When it comes to building new neighbourhoods, there are two options. We can build around roads or build around rail.
Building around roads is a somewhat underrated idea in Ireland (in that it’s so unpopular). But building around roads is a perfectly reasonable strategy. It can work. Cities of ten million people are built around roads and they work pretty well.
But we’d have to really commit to the idea. We’d need to bulldoze neighbourhoods to build giant 12 lane tolled motorways. We’d need ring roads upon ring roads. Suburban retail parks. Drive through banks. The whole nine yards.
I don’t think this is the path Ireland wants to follow. Irish people don’t love their cars as much as Texans. We have our climate targets. Successive governments wouldn’t make the investments.
The other option is to build around rail. Rail has some nice features: it has a huge carrying capacity (about 50x the equivalent width of roadway), so it doesn’t take up much space. It naturally lends itself to pleasant, walkable, sustainable, mixed use urbanism. Builders like it because it enables very dense developments. Rail companies like lots of buildings near their lines because they provide passengers for the trains.
The problem with rail is that it’s hard to build. It takes a lot of coordination, ambition, foresight, and technical skill. I don’t need to labour this point. Metrolink’s 12 kilometres of tunnel is nearly 20 years in the making.
There is a way, though, to get much of the benefits of a metro system with much less effort, investment and risk. It could unlock a giant city-spanning train network with the capacity of a metro, some 78 kilometres in length. The network would enable much greater density within the city and open up 15 kilometres of green fields, on which some 135,000 houses could be built right away.
The main thing we’d need to build is one short tunnel, 3.5 kilometres in length.
An S-bahn
This is an idea borrowed from 1930s Germany. Berlin at the time was congested. It sat at the centre of a spider's web of rail lines. But the rail lines weren’t designed to connect the city of Berlin; they were designed to connect the hinterland to the city of Berlin. The rail lines terminated at the edge of the city at northern, southern, eastern and western termini.
Berlin’s idea was to connect the northern and southern rail termini with a tunnel. This had three main benefits. First, it allowed travellers from the north to go to destinations in the south and vice versa. Second, it created new stations in the centre of the city, along the tunnel’s route. Third, it allowed trains to run at a much greater frequency, because they didn't need to be turned around at the city terminus. These things came to be known as Stadtschnellbahns, or S-bahns for short.
This idea worked. It was copied by: Bremen, Dresden, Hamburg, Hanover, Magdeburg, Leipzig-Halle, Munich, Nuremberg, Frankfurt, Mannheim, the Rhein-Ruhr Metropolitan Region, Rostock, Stuttgart, Vienna, Zurich, Milan, Stockholm, Prague, Copenhagen, Rotterdam, Paris, Stockholm, London and Melbourne. Among other cities.
The attached graphic (in German, but decipherable) does a good job of depicting S-bahns’ strengths and weaknesses. They are good at connecting the city to the hinterland, but not as good at connecting locations within the city as a normal metro system. They also require much less tunnelling and thus, are cheaper to build.
Also attached is a map is of Munich’s commuter rail system. Note the thick trunk at the centre of the network. This is the tunnel. Almost all lines converge on it. It connects the big stations at the west and east of the city, and runs through the city centre.
S-bahns make sense in Europe because European cities have a legacy of heavy rail. Where once the cities needed trains to deliver raw materials and goods, now they need trains to move people around the metropolitan area. Modern S-bahns repurpose industrial rail as electrified, high capacity people movers.
Dublin is a fine candidate for the S-bahn treatment.
A plan
We should build a 3.5 kilometre tunnel linking Connolly and Heuston. The tunnel would let people travel from the west to the east, the east to the west, the west to the centre, and the east to the centre.
By segregating this system from intercity services, giving it dedicated lines, we could massively increase its capacity. (This is the problem with the existing Phoenix Park tunnel – it is shared with intercity services, which greatly limits its capacity.)
Capacity matters because, circling back to the beginning, we need places to build houses. When you’re seeking to build say 200,000 homes in Dublin over ten years, the bottleneck is transport capacity. The greater the capacity of the commuter rail system, the more homes can be built.
Irish transit heads will recognise this idea. It’s a first cousin of the DART+ project, which was considered at length before being shelved in 2021.
Besides all the standard arguments for building an S-bahn in a congested city, in Dublin, there’s one further sweetener: there are 15 kilometres of empty green fields directly west of Dublin, between Adamstown and Sallins. That stretch of land is Dublin’s single best opportunity to alleviate its housing shortage.
How many homes could be built there? The idea is to build homes within a five minute walk of the station, so that rail transport is genuinely convenient. That’s about an 800 metre radius, which comes to 201 hectares. I interviewed an expert on Transit Oriented Development, Professor John Renne, about this in a piece in The Currency. Prof Renne said “To justify having a train station, you need a minimum of 4,000 households [within 800 metres of the station]. On the medium side, it’s around 12,000-15,000. And on the maximum side, you’re talking about maybe a central downtown location or a really intense area, you would be looking at upwards of 25 to 30,000 units within 800 metres.”
For west Dublin, let’s assume 15,000 homes within 800 metres of the station. Along the 15 kilometre line, there’s room for nine stations. That means there’s room for about 135,000 homes within walking distance of a station.
The goal would be for this to be a new business and cultural quarter of the city, as opposed to simply a dormitory suburb. It should be a place to live, work, shop and relax. Both because nice places are nice in themselves, and because it’s more efficient for the trains to be full in each direction throughout the day. We don't want everyone going east in the morning and west in the evening. Professor Renne said: “Generally, you need a really strong employment base. So it’s not just the 12 or 15,000, or 20,000 households near the station, but it gets into the number of jobs nearby also. You’d probably want to see at least half that number — or more — of jobs.”
135,000 homes in a new quarter of Dublin would be a good start. But building the tunnel, and the network, would enable so much more housing than that. The capacity and usefulness of the entire 78 kilometre DART network would be massively upgraded, along with the potential for new housing along those lines. A bigger more capacious network would allow the whole city to densify naturally.
To be sure, a Dublin S-bahn would not be a simple project. The final five kilometres of rail leading to Heuston Station are currently single tracked; they would need to be widened. The whole line would need to be electrified. To keep the network segregated from intercity lines (and thus, unlock more capacity) the trains to Wicklow and Wexford might get the chop. Either that or the track would need to be widened the whole way through south Dublin. Building out roads, water, schools and so on for this new city quarter would not be trivial.
But these problems are solvable. We have started to build the tools we’d need for this job. Urban development zones, which are part of the new planning act, are a perfect tool for masterplanning new neighbourhoods. If we paired UDZs with land readjustment, we could get the landowners on board.
Land value uplift along the Adamstown–Sallins corridor could pay for much of the tunnelling and other capex. Assuming €12 million per hectare uplift on net-developable land, 80 per cent of gross catchments redevelopable, and a 33 per cent capture rate, the discounted value the could be captured by the state is around €4 billion. (How to capture the value? Again, land readjustment).
However one looks at it — number of homes unlocked, cost efficiency, carbon efficiency, pleasantness of new places, cost to exchequer, commute minimisation — this would seem to be the best investment we could make in the future of the city.
And when we’re done we can do something similar in Drogheda, Limerick Junction, Cork, Limerick, Galway and Waterford.
3
89
Chris Heather retweeted
A place to put the houses
You are likely familiar with our national housing targets. The latest is 50,500 homes per year.
It’s hard to put this number in context. What do 50,500 homes look like? Is it too many? Too few?
One answer is that 50,500 homes look like Cork City. We need to build one Cork City this year, and next year, and the following year, for a couple of decades. Another answer is that it looks like building a new Dublin City every four years.
The construction of 50,500 homes per year isn’t something that can be achieved by building here and there, on odd plots of land. Building that many homes demands massive ambition, planning and forward investment.
Take City Edge, for example. City Edge is (by Irish standards) an ambitious plan to turn 300 hectares of industrial land in West Dublin into high density masterplanned housing. The City Edge plan is expected to take 45 years. By 2070, it’ll have contributed 40,000 homes in total. That’s about 900 homes per year, or 1.7 per cent of our annual target. Or take Cherrywood. Cherrywood is the biggest Strategic Development Zone in the history of the state. It’s ten years in the making. When completed in the late 2020s, it’ll have yielded… 10,500 homes.
All of which is to say, we need a bigger plan for where we’re going to put all these houses.
The two paths
When it comes to building new neighbourhoods, there are two options. We can build around roads or build around rail.
Building around roads is a somewhat underrated idea in Ireland (in that it’s so unpopular). But building around roads is a perfectly reasonable strategy. It can work. Cities of ten million people are built around roads and they work pretty well.
But we’d have to really commit to the idea. We’d need to bulldoze neighbourhoods to build giant 12 lane tolled motorways. We’d need ring roads upon ring roads. Suburban retail parks. Drive through banks. The whole nine yards.
I don’t think this is the path Ireland wants to follow. Irish people don’t love their cars as much as Texans. We have our climate targets. Successive governments wouldn’t make the investments.
The other option is to build around rail. Rail has some nice features: it has a huge carrying capacity (about 50x the equivalent width of roadway), so it doesn’t take up much space. It naturally lends itself to pleasant, walkable, sustainable, mixed use urbanism. Builders like it because it enables very dense developments. Rail companies like lots of buildings near their lines because they provide passengers for the trains.
The problem with rail is that it’s hard to build. It takes a lot of coordination, ambition, foresight, and technical skill. I don’t need to labour this point. Metrolink’s 12 kilometres of tunnel is nearly 20 years in the making.
There is a way, though, to get much of the benefits of a metro system with much less effort, investment and risk. It could unlock a giant city-spanning train network with the capacity of a metro, some 78 kilometres in length. The network would enable much greater density within the city and open up 15 kilometres of green fields, on which some 135,000 houses could be built right away.
The main thing we’d need to build is one short tunnel, 3.5 kilometres in length.
An S-bahn
This is an idea borrowed from 1930s Germany. Berlin at the time was congested. It sat at the centre of a spider's web of rail lines. But the rail lines weren’t designed to connect the city of Berlin; they were designed to connect the hinterland to the city of Berlin. The rail lines terminated at the edge of the city at northern, southern, eastern and western termini.
Berlin’s idea was to connect the northern and southern rail termini with a tunnel. This had three main benefits. First, it allowed travellers from the north to go to destinations in the south and vice versa. Second, it created new stations in the centre of the city, along the tunnel’s route. Third, it allowed trains to run at a much greater frequency, because they didn't need to be turned around at the city terminus. These things came to be known as Stadtschnellbahns, or S-bahns for short.
This idea worked. It was copied by: Bremen, Dresden, Hamburg, Hanover, Magdeburg, Leipzig-Halle, Munich, Nuremberg, Frankfurt, Mannheim, the Rhein-Ruhr Metropolitan Region, Rostock, Stuttgart, Vienna, Zurich, Milan, Stockholm, Prague, Copenhagen, Rotterdam, Paris, Stockholm, London and Melbourne. Among other cities.
The attached graphic (in German, but decipherable) does a good job of depicting S-bahns’ strengths and weaknesses. They are good at connecting the city to the hinterland, but not as good at connecting locations within the city as a normal metro system. They also require much less tunnelling and thus, are cheaper to build.
Also attached is a map is of Munich’s commuter rail system. Note the thick trunk at the centre of the network. This is the tunnel. Almost all lines converge on it. It connects the big stations at the west and east of the city, and runs through the city centre.
S-bahns make sense in Europe because European cities have a legacy of heavy rail. Where once the cities needed trains to deliver raw materials and goods, now they need trains to move people around the metropolitan area. Modern S-bahns repurpose industrial rail as electrified, high capacity people movers.
Dublin is a fine candidate for the S-bahn treatment.
A plan
We should build a 3.5 kilometre tunnel linking Connolly and Heuston. The tunnel would let people travel from the west to the east, the east to the west, the west to the centre, and the east to the centre.
By segregating this system from intercity services, giving it dedicated lines, we could massively increase its capacity. (This is the problem with the existing Phoenix Park tunnel – it is shared with intercity services, which greatly limits its capacity.)
Capacity matters because, circling back to the beginning, we need places to build houses. When you’re seeking to build say 200,000 homes in Dublin over ten years, the bottleneck is transport capacity. The greater the capacity of the commuter rail system, the more homes can be built.
Irish transit heads will recognise this idea. It’s a first cousin of the DART+ project, which was considered at length before being shelved in 2021.
Besides all the standard arguments for building an S-bahn in a congested city, in Dublin, there’s one further sweetener: there are 15 kilometres of empty green fields directly west of Dublin, between Adamstown and Sallins. That stretch of land is Dublin’s single best opportunity to alleviate its housing shortage.
How many homes could be built there? The idea is to build homes within a five minute walk of the station, so that rail transport is genuinely convenient. That’s about an 800 metre radius, which comes to 201 hectares. I interviewed an expert on Transit Oriented Development, Professor John Renne, about this in a piece in The Currency. Prof Renne said “To justify having a train station, you need a minimum of 4,000 households [within 800 metres of the station]. On the medium side, it’s around 12,000-15,000. And on the maximum side, you’re talking about maybe a central downtown location or a really intense area, you would be looking at upwards of 25 to 30,000 units within 800 metres.”
For west Dublin, let’s assume 15,000 homes within 800 metres of the station. Along the 15 kilometre line, there’s room for nine stations. That means there’s room for about 135,000 homes within walking distance of a station.
The goal would be for this to be a new business and cultural quarter of the city, as opposed to simply a dormitory suburb. It should be a place to live, work, shop and relax. Both because nice places are nice in themselves, and because it’s more efficient for the trains to be full in each direction throughout the day. We don't want everyone going east in the morning and west in the evening. Professor Renne said: “Generally, you need a really strong employment base. So it’s not just the 12 or 15,000, or 20,000 households near the station, but it gets into the number of jobs nearby also. You’d probably want to see at least half that number — or more — of jobs.”
135,000 homes in a new quarter of Dublin would be a good start. But building the tunnel, and the network, would enable so much more housing than that. The capacity and usefulness of the entire 78 kilometre DART network would be massively upgraded, along with the potential for new housing along those lines. A bigger more capacious network would allow the whole city to densify naturally.
To be sure, a Dublin S-bahn would not be a simple project. The final five kilometres of rail leading to Heuston Station are currently single tracked; they would need to be widened. The whole line would need to be electrified. To keep the network segregated from intercity lines (and thus, unlock more capacity) the trains to Wicklow and Wexford might get the chop. Either that or the track would need to be widened the whole way through south Dublin. Building out roads, water, schools and so on for this new city quarter would not be trivial.
But these problems are solvable. We have started to build the tools we’d need for this job. Urban development zones, which are part of the new planning act, are a perfect tool for masterplanning new neighbourhoods. If we paired UDZs with land readjustment, we could get the landowners on board.
Land value uplift along the Adamstown–Sallins corridor could pay for much of the tunnelling and other capex. Assuming €12 million per hectare uplift on net-developable land, 80 per cent of gross catchments redevelopable, and a 33 per cent capture rate, the discounted value the could be captured by the state is around €4 billion. (How to capture the value? Again, land readjustment).
However one looks at it — number of homes unlocked, cost efficiency, carbon efficiency, pleasantness of new places, cost to exchequer, commute minimisation — this would seem to be the best investment we could make in the future of the city.
And when we’re done we can do something similar in Drogheda, Limerick Junction, Cork, Limerick, Galway and Waterford.
Chris Heather retweeted
I wrote a piece for the @IrishTimes on why Ireland can’t build anything these days, and how to fix that.
John Collison of Stripe: Ireland is going backwards. Here’s how to get it moving irishtimes.com/life-style/pe…
Chris Heather retweeted
You wanted a nice day out flying kites with your family. But @NaturalEngland won't let you.
Natural England won't let us build, won't let us have fun. It won't let us do anything.
The Government empowers them to block everything. This has to end.
Ireland .... Ireland ... Ireland 🙄
We used to have average power prices that are much more similar between European countries before.
In 2018, almost all between 40 and 60 €/MWh.
IN 2024, variation from 25 to 100+ €/MWh.
Pick your spot carefully for reindustrialization.
(PS: color scales are different).
81
Worth a read . Ireland has very similar issues
Foundations: Why Britain Has Stagnated.
A new essay by @bswud, @SCP_Hughes & me.
Why the UK's ban on investment in housing, infrastructure and energy is not just a problem. It is *the* problem.
And how fixing it is the defining task of our generation.
ukfoundations.co/
2
166
Chris Heather retweeted
“You get books out of the machines”
The best response to degrowth is to share Hans Rosling’s experience of washing machines
This is a massive government failing . What are the @greenparty_ie even doing in government? Surely this is how they move the needle on climate
Reading through today's Draghi report on the future of Europe's economy. A piece which won't surprise, but should concern an Irish audience, on permitting for renewables Ireland called out for having amongst the longest timelines for both solar and wind.
71
Chris Heather retweeted
I watched this 100 times 😂😂😂🔥🔥
xitter.gq/MadisonKittay/status/1…
Fantastic interview .. PLAY ON..💪
If you have a few moments to listen to this interview as a parent with your child.
Possible one of the best post race interviews I have ever watched #PlayOn
3
73
Rickie 💀
Woke up thinking about where Xander's major season ranks among the best majors seasons...
His 2024 major finishes read 8th-1st-7th-1st.
Here's the best "Win-Loss-Tie Percentages" in the last 40 years
(𝘦𝘴𝘴𝘦𝘯𝘵𝘪𝘢𝘭𝘭𝘺, 𝘸𝘩𝘢𝘵 𝘱𝘦𝘳𝘤𝘦𝘯𝘵𝘢𝘨𝘦 𝘰𝘧 𝘵𝘩𝘦 𝘮𝘢𝘫𝘰𝘳 𝘧𝘪𝘦𝘭𝘥𝘴 𝘪𝘯 𝘢 𝘴𝘦𝘢𝘴𝘰𝘯 𝘥𝘪𝘥 𝘺𝘰𝘶 𝘣𝘦𝘢𝘵. 𝘞-𝘓-𝘛 𝘮𝘦𝘢𝘯𝘴 𝘱𝘭𝘢𝘺𝘦𝘳𝘴 𝘣𝘦𝘢𝘵-𝘱𝘭𝘢𝘺𝘦𝘳𝘴 𝘭𝘰𝘴𝘵 𝘵𝘰-𝘱𝘭𝘢𝘺𝘦𝘥 𝘵𝘪𝘦𝘥)
99.3% Tiger 2000 (548-4-0) 🏆🏆🏆
99.1% Spieth 2015 (556-4-1) 🏆🏆
99.1% Tiger 2005 (552-4-1) 🏆🏆
98.5% Brooks 2019 (543-5-3) 🏆
98.2% Phil 2004 (547-8-2) 🏆
97.9% Rickie 2014 (549-8-4)
𝟵𝟳.𝟰% 𝗫𝗮𝗻𝗱𝗲𝗿 𝟮𝟬𝟮𝟰 (𝟱𝟭𝟲-13-1) 🏆🏆
97.2% Norman 1986 (528-13-2) 🏆
97.2% Rory 2022 (513-14-1)
97.0% Ernie 2004 (540-13-4)
95.8% Rahm 2021 (529-13-10) 🏆
Statistically, Xander's 2024 is the 7th best major season in the last 40 years.
54
