A.I. generativa o Stupidità Artificiale generalizzata? | Anne Alombert | TEDxLakeComo
The speaker argues against the over-anthropomorphization of AI by emphasizing that current generative tools only mimic language through statistical probability, not genuine thought. She contends that the greater danger is the "progressive destruction of our individual and collective minds" due to the "attention economy," rather than an artificial superintelligence. To counteract this, the speaker recommends developing technologies that promote collective deliberation and critical thinking, such as Wikipedia.
## Speakers & Context
- Speaker addresses the audience, stating she will speak in French as well.
- Context involves discussing the ubiquitous concept of Artificial Intelligence (A.I.) and its implications on human cognition.
## Theses & Positions
- The danger posed by AI is not the advent of artificial superintelligence but the "progressive destruction of our individual and collective minds," which constitute a new resource for digital industries.
- AI tools like ChatGPT and Midjourney generate content based on "statistical calculations made on huge amounts of data," simulating ability rather than possessing it.
- The core threat lies within the "attention economy," which uses "persuasive technologies" to capture attention for the purpose of selling data through targeted ads.
- Over-reliance on digital media risks causing "proletarianisation," defined as the "loss of our skills and capabilities due to their being delegated to technical devices."
- The true goal of digital technologies should be to serve "collective intelligence," not to automate thought processes.
## Concepts & Definitions
- **Anthropomorphism:** The tendency to attribute human characteristics, such as inner life, emotions, or conscience, to non-human entities like machines.
- **Transhumanist theories:** The ideology underlying major digital companies, advocating for technological enhancement of human capabilities.
- **Deep attention:** The ability to concentrate for long periods on a single object or activity, characteristic of print media.
- **Hyper attention:** A divided and fragmented attention, dispersed across multiple contemporary activities, characteristic of digital media.
- **Attention economy:** A business model aiming to capture human attention for as long as possible through persuasive technologies.
- **Captology:** A technoscience blending computer science, algorithmics, design, behavioral neurosciences, and cognitive psychology to develop technologies that influence thought and behavior.
- **Pharmakon:** A Greek term used by Plato meaning both remedy and poison, illustrating something that can have dual, contradictory effects (e.g., writing).
- **Artificial stupidity:** A concept describing the potential for technologies to diminish human cognitive capacities.
- **Proletarianisation:** The loss of skills and capabilities when they are delegated to technical devices.
## Mechanisms & Processes
- **Generative AI Function:** Creates text or images through statistical calculations on massive datasets, predicting the most probable word or sequence of letters in real time.
- **Attention Capture Mechanisms:** Include constantly sending notifications or implementing "infinite scrolling," which prevents users from deciding what to engage with.
- **Dopamine Stimulation:** Features like quantifying visits stimulate the brain's reward system, causing a slightly compulsive return to social networks.
- **Algorithmic Profiling:** Data collection and algorithmic calculations allow for the creation of standardized user profiles for personalized content targeting.
- **Cognitive Skill Degradation:** The shift from deep to hyper attention, and the potential abandonment of reading/writing, can impact "brain activity and on certain mental and social behaviours."
## Timeline & Sequence
- **Ancient Greece (Plato's "Phaedrus"):** Initial reflection on writing, identifying its dual nature as both remedy for memory and poison to memory effort.
- **2007:** Katherine Hayles observed the shift from print to digital media impacting attention regimes.
- **Recent times:** Highlighted by studies concerning the impact of social networks (e.g., "Facebook Files" on Instagram) and algorithmic feeds (e.g., TikTok's "Deadly by design").
## Named Entities
- **Raymond Kurzweil:** Entrepreneur, initiator of the transhumanist movement, former Google Director of Engineering, who discusses "intelligent machines" or "spiritual machines."
- **Yann Le Cun:** AI specialist at Facebook, who speaks of "learning machines."
- **Elon Musk:** Co-signer of a recent letter warning about AI dangers, using the term "digital minds."
- **Michel Desmurget:** Neuroscientist in France arguing that excessive screen time harms concentration, memory, and language development in youngsters.
- **Katherine Hayles:** American researcher who noted the shift in "regime of attention" from print to digital media.
- **Maryanne Wolf:** American researcher specializing in neuroscience and psycholinguistics, who stresses reading's role in empathy and synapse formation.
- **Frances Haugan:** Former Facebook employee who revealed the "Facebook Files."
- **Center for Countering Digital Hate:** Issued a report titled "Deadly by design."
- **Bernard Stiegler:** Philosopher who described the risk of "artificial stupidity" or "generalized proletarianisation."
- **Wikipedia:** Mentioned as an example of a collaborative encyclopedia supporting collective deliberation.
- **Tournesol association:** Suggested implementing a "collaborative recommendation" system.
- **Mastodon:** Social network example allowing users/groups to set content recommendation parameters.
## Numbers & Data
- Decade difference between the decline in manual skill and the Industrial Revolution was already present (comparison point).
- Plato's discussion of writing is detailed in his work *"Phaedrus."*
## Examples & Cases
- **AI Simulation:** ChatGPT and Midjourney generate content using statistical probability, predicting the most probable word sequence.
- **Neuroscientific Impact:** Maryanne Wolf notes that adopting a character's point of view when reading/watching is how learning builds empathy.
- **Social Media Harm:** Instagram's impact on teen mental health ("Facebook Files"); TikTok's algorithm luring vulnerable teens into risky information loops ("Deadly by design").
- **Algorithmic Suggestion:** Predictive text software suggests commonly searched terms, leading users to conform to the majority.
- **Proletarianization Parallel:** Craftsmen losing know-how to machine tools is paralleled with citizens risking loss of intellectual capability through algorithms.
## Tools, Tech & Products
- **Smart objects, smart phones, smart cities:** Ubiquitous examples of technology discussed.
- **Generative AIs:** Tools like ChatGPT and Midjourney.
- **Predictive text software:** Uses algorithmic calculations to suggest the next word.
- **Infinite scrolling:** An IT technology mechanism used to maximize time on the platform.
- **Social networks:** Platforms like Instagram and Mastodon.
- **Collaborative Encyclopedia:** Wikipedia.
- **Recommendation Systems:** Both commercial (digital giants) and alternative (Tournesol/Mastodon).
## References Cited
- Plato: *Phaedrus*.
- Center for Countering Digital Hate: Report titled "Deadly by design."
## Trade-offs & Alternatives
- **Print media (Book/Newspaper):** Offers "deep attention" (concentration on one object).
- **Digital media:** Demands "hyper attention" (divided, fragmented focus across many activities).
- **Digital Content Creation:** Potential benefit of massive data storage/symbol generation vs. the risk of destroying individual/collective decision-making skills.
- **Technical Solution:** Developing and experimenting with technologies that actively train memory, reflection, interpretation, and collective deliberation skills.
## Counterarguments & Caveats
- The discourse of AI's dangers often focuses on "intelligence" or "spirituality," diverting attention from documented harms related to screen time and mental skills.
- The concept of "thinking" or "creative" machines is flawed because human invention involves *disrupting* averages, not calculating them.
## Methodology
- Comparison of historical technological shifts (writing, machine tools) to modern digital technology development.
- Analysis of the underlying business model of the "attention economy" to explain technological design.
- Recommendation for an alternative model based on **collective deliberation** rather than individual optimization.
## Conclusions & Recommendations
- Stop debating the myth of artificial intelligence's consciousness.
- Instead, design and develop technologies that facilitate human cognitive skills—memory, reflection, interpretation, and collective deliberation.
- Implement decentralized models like Mastodon or collaborative recommendation systems like those proposed by Tournesol to counteract the interests of "digital giants."
## Implications & Consequences
- The underlying consequence of the attention economy is the commodification of human attention, leading to the erosion of mental skills.
- Misinformation spreads because the infrastructure captures attention and bypasses deep, critical thought.
- The stakes are the loss of human capacity for original thought, which becomes a resource to be exploited.
## Verbatim Moments
- *"what is the question that you are left with? What is it that you want to know?"* (Implied rhetorical moment leading to the core question).
- *"We are constantly glued to our screens, even at night."*
- *"The aim is indeed to keep us connected and imprisoned as much as possible, in order to sell our data, or our attention, to companies that will bombard us with targeted ads."*
- *"The invention does not consist in calculating averages but rather in disrupting averages and habits of thought in order to create 'the infinitely improbable' against 'the laws of statistics and their probability'."*
- *"And just as artisans had become proletarian workers, citizens risk becoming proletarian users."*
- *"what seems to improve us, can also diminish us."*
- *"This is what seems to be happening with the latest forms of contemporary digital and reticular writing."*
- *"In other words, the time has come to ditch the myth of artificial intelligence, and put digital technologies at the service of collective intelligence."*