Notesale: Turn your study into money

Document Preview

Extracts from the notes are below, to see the PDF you'll receive please use the links above

---

Platform Capitalism and Automation
Part 1 - Platform Capitalism
Section 1 - What are platforms?
Platform Capitalism

• Part of a new lexicon within economic geography - e
...
gig economy

• Most (in)famous examples - UBER, Deliveroo, Airbnb

• Issues with workers’ rights, gentrification (Berlin has banned Airbnb, Uber’s
licence in London was revoked)

Histories of platforms

• Idea is both new and old

• Extends lean, digital and knowledge economies - simplifying a company (e
...

only designing a product and not owning factories, but outsourcing production)

• Key differences exist - companies like Amazon own many parts of the production
process

• Etymology of the term is important - neutral and modern - means people perhaps
don’t question its worth

• Term was coined by bloggers, lacks clarity

• Shows how academia often happens behind the curtains, took longer to catch
up

• ‘Black box’ economy intentionally opaque (Gillespie, 2017) - people don’t look
into it if they don’t understand it

Defining the platform

• ‘Multi-sided markets’ (Langley and Leyshon, 2016)

• Digital infrastructures enabling two or more groups to interact

• Two types of platform:

• 2 sided - e
...
Apple, producer of apps and consumer

• 3 sided - e
...
Android, platform in the background, producers of apps (can
be anyone) and consumer

Platform and data

• Data, the central part of platforms

• New ‘raw material’ to mine, monopolise, and analyse (Srnieck, 2017)

• ‘Big data’

• Platforms’ network effects drive this - the more people that use the platform, the
more data can be collected - more people will use it through advertising to target
markets through data collection

• Key for knowledge economy and high value industry

• Data needs cleaning and analysis

• Small, highly-skilled workforce

• Data storage expensive

Platforms and ‘network effects’

• “platforms are not simply in the business of intermediating connections, but of actively
curating connectivity” (Langley and Leyshon, 2016)

• The more people who use a platform, more useful it becomes - e
...
Google is so
much more useful than other search engines because so many people use it

• And more data platform owners access


Free platforms?

• Some platforms cost - e
...
Spotify

• Data improves service, gets more customers

• Also drives targeted advertising

• Other platforms ‘free’ - e
...
Google, Facebook

• ‘Cost’ = your data

• ‘Free’ elements ‘cross-subsidise’ (Google)

• Data monopoly gives platforms ‘rent’

• Network effects

• 5 types of platform (Srnicek, 2017)

Advertising platforms

• Google (89%) and Facebook (96
...


• ‘Open system’, meaning this has led to ad-blocking, costing Google $27 billion in
2014

• Facebook

• Hyper-detailed profiling (algorithms), increasingly closed system to prevent adblocking

• Rise in mergers and monopolisation - Google buying one company a week, anything
which looks like it might provide more data

Cloud platforms

• Amazon web services lease out hardware, software, apps

• Lowers barriers to entry

• Easy for companies to ‘scale’

• AWS most profitable arm of company

• Worth $70 billion, more than all other selling parts

• Amazon loses money on Prime/Kindles - but they do it because it locks you in to
their system to gather data on you

• Gives Amazon access to all data to modify company

• Beyond ‘lean production’ and ‘core competencies’

• Increasingly hard to run a website without Amazon web services

Industrial platforms

• Industrial ‘internet of things’

• Sensors and automation = hyper-lean, flexible, ‘just in time’ production production lines can now produce a variety of things because sensors on every
aspect of the process means changes can be made on the way

• ‘Technical fix’ reduces production costs

• Labour - 25%, energy - 20%, wear and tear - 40%

• Huge amounts of data produced through sensors

• Improves production and monitoring

• Massive fixed capital investment means that huge amounts of money are require
in the first place, resulting in an…

• Oligopoly - e
...
Siemens and GE - calls into question the assumption that leaving
things to the free market (rather than state involvement) will prevent monopolies/
oligopolies forming


Product Platforms

• Involves leasing out a product

• Important with stagnating wages - no ‘big ticket’ purchases

• Music streaming is a prime case

• Usability and price is key - Spotify vs Sky Sports (Spotify is cheaper, more
people therefore use it)

• Rolls-Royce makes most of its money from jet engines - ‘lease thrust’

• Higher margins in repair, retain performance data through sensors

• Data fed-back for competitive advantage

• Network effects and oligopoly

Lean platforms

• Most well-known and controversial platform type - Airbnb, UBER, etc
...
g
...
g
...
1 million lost, 2
...
g
...
g
...
02g CO2, as servers have to be fired up

Geographies of platforms

• Extending ‘NIDL’ (Graham et al, 2017)

• Tasks can be done anywhere in the world

• Provides work in the global south, avoids isolation

• Workers competing and undercutting

• Improves logistics, GPNs

• Command and control areas retain power

• High-tech clusters

• Lends itself to tax evasion

The gig economy

• Platform that allows connection between buyer and seller - an intermediary

• Workers pick up ‘gigs’ at own behest

• Affords flexibility to worker…

• …but saves company up to 30% as avoiding employer duties

• Much job-growth in this sector

• Emerging at a time of austerity and unemployment - people taking any job they
could get

• Who can work and live like that? - favours single people, rather than those with
families

Deliveroo

• Workers ‘self-employed’ contractors, however:

1
...
Be available for certain shifts

3
...
75 per drop, rather than £7 p/h plus £1 p/d as it used to be

• Even less security, favours certain riders - faster riders, those in busier areas

• Realistic full-time employment?

• Monitoring of work by app

• ‘Algorithmic management’

• Gives workers the fastest route, if this is not taken this is fed back to Deliveroo
who can discipline the workers

Uber

• Affective labour - workers rated by consumers, policing behaviour


•
•

Inbuilt racism and sexism - technology is never neutral - names which sound of
Muslim or Black origin are less likely to be picked up, women using Uber get asked for
longer routes as people assume they cannot use apps

Aggressive business practices

• Phantom cabs - filling up cabs by booking them then cancelling so people have
to use Uber instead

• Forcing competitors out - network effect

• Deskilling and wage suppression

• Legislation avoidance

• Will it come back to haunt them? - Uber still hasn’t made profit


New forms of organising and resistance

• Emergence of new unionism

• Rebel Roo

• Uber strikes

• Wider trend in trade unionism - people have so much downtime they meet up
and discuss working conditions

• Companies losing high-profile court cases

• Lean advantage disappearing?

• In-built risk that companies are prepared to take due to ability to move to
different cities as asset-less?

• Taylor review of modern working practices

Section 3 - The Future of Platforms
• The future of (lean) platforms

• Inherently vulnerable

• Most fail – need low skill & regularity

• Less efficient than alternatives (e
...
Airbnb vs a hotel)

• Clamping down on evasion - Uber

• ‘Growth before profit’ & VC bubble

• Uber trialling self-driving cars

• Developing own mapping software

• Technical fix

• Key growth area, first-mover advantage
• “outlet for surplus capital in an era of ultra-low interest rate and dire investment
opportunities [rather] than the vanguard destined to revive capitalism” (Srnicek, 2017)
The future of (industrial and cloud) platforms

• Monopoly/oligopoly

• ‘Platform wars’

• Networks effects and first-mover

• Controlling all of GPN?

• Automation and technical fixes

• Machine learning and AI

• Also digital-spatial fix

• Rates of profit still fall, over-production remains

The future of (advertising) platforms

• Home internet of things and wearables

• Expansion of data extraction is key

• Ecosystems closing


•

• App-based internet and ad-blocking

Increased mergers and cross-subsidisation

• Buying WhatsApp, LinkedIn


The future of (product) platforms

• Wage stagnation reduces major purchases

• Companies make more money from rental

• Economies of affect and ease

• Ratio of price and convenience

• Intimately linked other platforms


Part 2 - Automation
Section 4 - Automation
Automation
• Long history, but new contours
• Term coined by Ford’s vice-president in 1948
• Second machine age’/’fourth industrial revolution’

• Self-acting machines now analyse & synthesise
• (Big) data key - platforms
• Attitudes often at two extremes (Blissell & Del Casino, 2017)

• ‘Dystopian angst’ vs ‘positive boosterism’
• Rapidly creating new (economic) geographies 

Why automation?

• Technology never neutral
• Reflects emergent social, economic, and political situation
• Technology ‘ambivalent’ (Richardson, 2016)

• Exploring capitalist automation
• Linked to tendencies from Lecture 2 (i
...
profit)
• Intensifies & extends ‘work’
• ‘Technical fix’ to reduce costs and discipline labour
• Creates new ‘spatial fixes’ (GPNs and logistics)
• But, always potentials for different use
• Improve efficiency/productivity for repetitive tasks
• Machines on assembly lines with overseers
• Humans required to correct errors/for more complex tasks
• Robots reducing human employment

• “Machines must be rebuilt to engage in new functions
...
They can learn from their
environment and respond to it
...
6 million
What automation?
• Mental task < 1 second’s thought = probable automation
• 49% of activities in jobs right now could be automated
• 5% of jobs = full automation                      

• 60% of jobs = 30% automation
• Manufacturing jobs most obvious, but also services
• Data entry & retail
• Robot lawyers, academics, surgeons

•
•

•

Key thing is routine and repetition
• ‘Codified’ knowledge
Automation’s impacts vary, depend on economy
• UK = 30% of jobs ‘high risk’ by 2030 - lower than other countries due to servicebased economy
• Germany = 35%
• US = 38%
• Global South = 85%
Education is key
• ≥ degree = 12%; ≤ GCSE = 46%

Automation and job creation?
• Computational advances created over 1500 new job types
• ATMs freed up bank workers for new roles and branches
• Birth of high-skill, tech jobs - e
...
data analysis for platforms
• Self-service checkouts and security
• Humans watching robots; robots watching humans
• But, new roles don’t produce mass employment
• Sectors of mass employment most vulnerable
• 140 million ‘cognitive’ jobs at high threat

Section 5 - New geographies of automation?
Consequences of automation

• Creating mass ‘surplus populations’
• Jobless recovery’ from crisis
• Driving ‘gig’ employment
• Global reserve army reducing wages
• Intersects with gender, race, and nationality
• Women = 5 jobs lost, 1 created; men = 3 lost, 1 created
• Low-paid and ‘low-skilled’ jobs most vulnerable
• Vary according to sector and geography
• Textile industry barely changed

• Electronics = Pearl River Delta 1
...
g
...
with a small
workforce means the workforce is powerful - fixed capital investment means not
easy to move, therefore a required workforce can exert influence

• Creates monopolies - conditions where only a few companies can benefit
• Amara’s law: “We tend to overestimate the effect of a technology in the short run and
underestimate the effect in the long run” (Woodward, 2017)

‘Utopian’ automation
• Automation has potential to produce more and work less
• Keynes estimated working week at 15 hours
• Potential to limit ‘undesirable’ work
• Lean production can prevent waste, meet growing needs
• Goods are cheaper, balance out wage decline
• Create work in emerging green economy
• Open-source revolutions in sharing economy
• Worker-owned platforms? - For example, a localised Uber or community owned
Airbnb
Making automation work for us
• Requires response to ‘surplus population’
• Re-training?
• Universal basic income?
• Everyone irrespective of income earns a flat rate

• Proponents on both left (people won’t starve, can avoid taking bad jobs) and
right (can get rid of benefits) – concerning?
• Increasingly popular – Swiss referendum
• Possible globally?
• Regulation and worker organisation
• Ambivalence of technology remains… whether it is goo for bad is down to the will of
people

Geographies of automation
• New geographies of ‘premature deindustrialisation’
• Multi scalar, but particularly in the Global South
• Production &distribution more global than ever
• But still localised through agglomeration
• High-tech command and control
• High-tech production sites
• Full impacts of platforms and ‘second machine age’ under-explored
• (Economic) geography well placed, but lagging 


Conclusions
• Emergence of ‘platform capitalism’
• Multi-sided markets with network effects
• Data new resource and source of profit
• Different platform types more stable long-term
• Automation
• Entering ‘second machine age’
• Significant impact on job-market, but there are limits
• Technology not neutral
• Creating new geographies, but research lacking



---