With the growing influence of artificial intelligence (AI) in our daily existence, it's imperative that we engage in discussions about its potential impact on society and our day-to-day experiences.
2. With the growing
influence of artificial
intelligence (AI) in our
daily existence, it's
imperative that we
engage in discussions
about its potential
impact on society and
our day-to-day
experiences.
4. Artificial intelligence is currently
exerting a significant influence on
society, and this influence is
expected to magnify as the
technology advances further.
However, it's important to
acknowledge that not all of these
effects are necessarily beneficial. To
shed light on this, we've compiled a
list of six disadvantages of artificial
intelligence that merit our
attention.
6. Artificial intelligence can exhibit faults in
various ways, underscoring the crucial need
for transparency. Issues can arise from
inaccurate or inadequately cleaned input
data, or even from the inadvertent selection
of biased data sets during model training by
data scientists and engineers. However, the
primary challenge lies in the absence of clear
visibility: the inability to understand why AI
is underperforming or, at times, the
unawareness that it is indeed performing
poorly. In contrast, conventional application
development incorporates quality assurance
and testing processes, supported by tools
that swiftly identify and rectify any bugs.
8. AI can create highly accurate profiles by
analyzing personal data, particularly location
history. Algorithms excel at recognizing
patterns and can predict future locations with
even greater precision when incorporating
data from friends and social contacts.
Sometimes, the risks of AI are underestimated.
Even if you believe you have nothing to hide,
sharing personal information may not sit well
with you. Think of it as not wanting to live in a
glass-walled house. Would you be comfortable
sharing your device's location history? What
about your teenage daughter's data? It's likely
you'd prefer not to have her location data,
including predictions, publicly disclosed.
Remember, information is power, and
relinquishing information means surrendering
power.
10. The impact of automation and AI on the
job market raises concerns about
potential unemployment and job shifts.
Experts have varying predictions, with
estimates suggesting that anywhere
from 3 to 14 percent of the global
workforce, ranging from 75 million to
375 million workers, may need to change
occupations and acquire new skills by
2030. This divergence in predictions
underscores the lack of consensus
among technology and business experts
regarding the future of the labor market,
making it challenging to determine the
exact number of jobs that will be
affected.
12. The growth of disinformation, already witnessed due to
artificial intelligence, is a significant concern. In 2020,
Extinction Rebellion used AI to create a deepfake,
altering a genuine video of Belgian Prime Minister
Sophie Wilmès to spread false information.
Deepfakes are expected to play a growing role in
targeted disinformation campaigns, jeopardizing
democratic processes and fostering societal division.
Additionally, online bots can generate fake content like
altered news articles and tweets to promote deceptive
agendas. For example, AI tool GPT-3 recently crafted
tweets aiming to sow doubt about climate change.
These technologies, coupled with figures like Trump
labeling the media as fake, could lead to what The
Atlantic termed the "collapse of reality." With deepfakes
and online bots disseminating disinformation, there's a
risk of blurred lines between reality and fiction, eroding
trust in our political institutions.
14. Big Tech companies, led by Google since 2007, have
acquired around 30 AI firms, establishing a significant
monopoly in AI technology. However, Google is not the
sole player in this scenario. In 2016, Google, Apple,
Facebook, Microsoft, and Amazon, along with major
Chinese tech giants, collectively invested approximately
$30 billion out of an estimated global total of $39 billion
in AI-related research, development, and acquisitions.
The widespread acquisition of AI startups globally raises
concerns as these tech giants wield significant influence
in shaping the direction of AI technology. With their
dominance in areas like search, social media, online
retail, and app stores, these companies hold near-
monopolies on user data and are emerging as primary AI
suppliers to the broader industry. This concentration of
power poses risks, including the potential for large tech
companies to influence democratically elected
governments.
16. While AI can benefit the environment, such as in
smart grids and sustainable cities, it also poses an
environmental drawback due to its energy-intensive
nature. A 2019 study highlighted the substantial
carbon footprint of a specific AI type, particularly
deep learning in natural language processing, driven
by the energy demands of its hardware. Experts note
that training a single AI model emits about 300,000
kg of CO2, equivalent to 125 round-trip flights from
NYC to Beijing or five times an average American
car's lifetime emissions. It's worth noting that
emissions aren't solely tied to model training; the
infrastructure supporting AI deployment, like data
centers and material transportation, also
significantly impacts the environment.