Who Gets Access to Abundant Intelligence?

Introduction

If advanced AI eventually makes high-quality cognitive help cheap and widely available, one of the most important questions will not be how intelligent the systems become. It will be who gets access to them.

The optimistic vision behind abundant intelligence is that millions or billions of people gain access to forms of cognitive support that were previously scarce: tutoring, translation, research assistance, legal guidance, medical information, planning help, software development support and scientific tools. But history suggests that transformative technologies do not automatically spread their benefits evenly. Railways, electricity, higher education, the internet and modern medicine all produced large gains while also creating new forms of concentration and inequality before wider diffusion.

The same tension sits at the centre of AI bloom. Cheap cognitive help could become a kind of public infrastructure that expands opportunity across society. Or it could become a strategic resource controlled by a small number of firms, states or wealthy institutions. The difference may shape whether AI produces broad human flourishing or a more unequal version of technological progress.

Why access matters as much as capability

Much discussion of frontier AI focuses on whether systems become more capable. But from a social perspective, access can matter almost as much as raw capability.

A highly capable model that is expensive, restricted or available only to large corporations does not create abundant intelligence for society as a whole. By contrast, a somewhat less capable system that is affordable, reliable and broadly accessible may have much larger real-world effects on education, entrepreneurship and social mobility.

The distinction already appears in other domains. Scientific knowledge is valuable partly because universities, libraries and journals spread it. Electricity transformed societies not because generators existed, but because power networks eventually reached homes, schools and businesses. The internet became civilisation-scale infrastructure because billions of people could connect to it.

AI may follow a similar pattern. The largest social gains could come not from the smartest laboratory system but from the extent to which useful cognitive assistance becomes available to ordinary people.

This matters especially because intelligence is a force multiplier. People who receive better explanations, better planning support, better access to expertise and better educational tools often become more capable in many areas simultaneously. A farmer, nurse, student, entrepreneur or local official equipped with reliable cognitive assistance can solve problems that previously required scarce specialist support.

The bloom case therefore depends partly on diffusion. Intelligence only becomes abundant if access becomes abundant.

The new bottleneck: compute, data and infrastructure

One reason access is contested is that modern AI depends on infrastructure that is unusually concentrated.

Frontier systems require vast computing resources, specialised chips, large-scale data centres, engineering talent and enormous capital investment. A relatively small number of companies control much of this stack. Governments increasingly view advanced AI infrastructure as strategically important in the same way they view energy systems, telecommunications networks or semiconductor supply chains.[Tony Blair Institute]institute.globalsovereignty in the age of ai strategic choices structural dependenciesTony Blair InstituteSovereignty in the Age of AI: Strategic Choices, Structural…19 Jan 2026 — An agenda for states to strengthen AI so…

This creates a tension at the heart of abundant intelligence.

On one side, AI systems can make knowledge and expertise dramatically cheaper to distribute. Once trained, a model can serve millions of users simultaneously. That creates the possibility of cognitive abundance.

On the other side, the infrastructure needed to build and operate the most advanced systems may naturally concentrate power. The same technologies that could democratise expertise can also centralise control over the tools that provide it.

Researchers examining AI compute governance note that computing power is unusually concentrated, measurable and controllable compared with many other technological resources. That makes it attractive as a policy lever, but also raises concerns about dependency and centralisation.[Frontiers Policy Labs]policylabs.frontiersin.orgFrontiers Policy LabsDesigning capability: institutions and the future of emerging…December 16, 2025 — by V Dhar — Emerging compute is…Published: December 16, 2025

In practical terms, this means the future of cognitive infrastructure may be shaped not only by model quality but also by who owns the data centres, chips, cloud platforms and deployment channels through which intelligence is delivered.

What concentration could look like

The concern is not simply that some companies become successful. Large-scale infrastructure often requires large-scale organisations.

The deeper concern is that a small number of actors could gain disproportionate influence over how knowledge, reasoning and cognitive assistance are distributed.

Several forms of concentration are frequently discussed:

• Compute concentration. A handful of firms and countries possess the largest AI computing clusters and semiconductor supply chains. Access to advanced compute increasingly shapes who can train frontier systems.[OECD.AI]oecd.aithe geopgraphy of ai compute mapping what is available and whereThe geography of AI compute: Mapping what is available…Oct 29, 2025 — Policymakers and industry leaders need to measure and map AI com…

The risk is not merely economic. It is also political and cultural. If cognitive infrastructure becomes as important as electricity or communications networks, control over that infrastructure may translate into influence over education, media, public administration and scientific research.

OECD analyses of AI competition and productivity increasingly emphasise that concentration in AI development, uneven adoption and unequal diffusion could widen existing social and economic gaps if not addressed through policy.[OECD]oecd.orgThe impact of Artificial Intelligence on productivity…by F Filippucci · 2024 · Cited by 161 — The paper discusses the concentrati…[OECD]oecd.orgcomponent 6AI-related competition concerns in downstream marketsNov 14, 2025 — This paper examines how the adoption of artificial intelligence (AI)…

A new cognitive divide

The most familiar inequality story is that some people have AI tools and others do not.

The more important divide may be subtler.

Access alone does not guarantee benefit. People also need the skills, infrastructure and institutional support required to use AI effectively.

UNESCO has increasingly described an emerging “AI divide” in which unequal access to technology, education and digital capability determines who benefits from AI systems.[UNESCO]unesco.orgAI literacy and the new Digital DivideAI literacy and the new Digital Divide - A Global Call for…Aug 6, 2024 — This divide represents the unequal access, benefits, an…

Several layers of inequality may emerge simultaneously:

Infrastructure inequality

Many regions still face unreliable electricity, limited internet access or weak computing infrastructure. If advanced AI becomes central to education, research and economic participation, these gaps become more consequential.

Recent analyses of global AI diffusion argue that access to power, data centres and connectivity may become as important as access to software itself.[Business Insider]businessinsider.comHowever, this rapid uptake is uneven and is exacerbating a global digital divide. High-income nations like the UAE (59.4% adoption), Sing…

Language inequality

Most frontier models perform best in high-resource languages, especially English. People working in less represented languages may receive weaker assistance, fewer educational benefits and less effective local adaptation. UNESCO has repeatedly warned that AI systems must become more responsive to local linguistic and cultural contexts if benefits are to spread broadly.[UNESCO]unesco.orgguidance generative ai education and researchGuidance for generative AI in education and research7 Sept 2023 — UNESCO's first global guidance on GenAI in education aims to supp…

Institutional inequality

Large companies can hire AI specialists, purchase premium systems and redesign workflows around AI adoption.

A small school, local council, charity or rural clinic often cannot.

If access to cognitive infrastructure becomes a major productivity advantage, institutional capacity differences could widen even when the technology itself is widely available.

Literacy inequality

The ability to work effectively with AI is becoming a skill in its own right.

Recent research suggests that productivity gains from generative AI vary substantially across users. People who are better at prompting, checking and integrating model outputs often gain much more value than those who are not. Importantly, some studies suggest that training and structured workflows can reduce these gaps.[arXiv]arxiv.orgGenerative AI and the Productivity Divide: Human-AI Complementarities in EducationMay 18, 2026…Published: May 18, 2026

The lesson is that abundant intelligence requires more than distributing software. It also requires helping people use it well.

Why governments increasingly treat AI as infrastructure

One response to these risks is to think of advanced AI less like a consumer app and more like infrastructure.

Governments already treat roads, electricity, water systems, schools and telecommunications as foundational capabilities that support participation in modern society. Some policymakers increasingly argue that reliable cognitive assistance may eventually deserve similar treatment.

This does not necessarily mean state ownership of AI systems. Rather, it means recognising that broad access has public value.

Several emerging policy approaches reflect this idea:

UNESCO’s work on AI in the public sector focuses on building institutional capacity so governments can use and govern AI in the public interest rather than remaining dependent on external actors.[UNESCO]unesco.orgAI for the Public SectorUNESCO's AI for the Public Sector programme strengthens the institutional capacity of governments worldwide…

The underlying question is whether societies want cognitive capability to function primarily as a private luxury good or as a broadly available public resource.

Open models, public models and the argument over openness

One of the most important disputes concerns openness.

Supporters of more open AI ecosystems argue that broad access requires alternatives to fully closed systems. Open-weight models, shared research and distributed innovation can lower barriers to entry, support local adaptation and reduce dependence on a few providers.

The argument is partly economic. If only a small number of firms control advanced systems, competition may weaken and users may have fewer choices.

It is also partly democratic. More open ecosystems can allow universities, smaller companies, nonprofits and governments to inspect, adapt and deploy systems for local needs.

However, openness creates trade-offs.

More capable models can also be misused. Open release may increase risks related to cybercrime, misinformation, biological threats or other harmful applications. Policymakers therefore face a difficult balancing problem: expanding access without making dangerous capabilities easier to exploit.

OECD work on AI openness highlights that the issue is not binary. Different systems can be open in different ways, with varying levels of access to weights, code, training details and deployment rights. The debate is increasingly about finding governance models that preserve innovation and broad participation without ignoring security concerns.[OECD]oecd.org7376c776 enEmerging divides in the transition to artificial intelligenceby S Kergroach · 2025 · Cited by 28 — It compares business adoption rates ov…

Public-interest safeguards for reliable cognitive help

If abundant intelligence is to become genuine cognitive infrastructure, reliability and accountability matter as much as availability.

People will increasingly rely on AI systems for decisions involving education, employment, finance, health and public services. Errors, biases and manipulation can therefore have large consequences.

Several safeguards are frequently proposed.

Independent evaluation

External testing can help determine how systems perform across different populations and use cases.

The UK’s AI Security Institute and similar organisations have argued that systematic evaluation is necessary because model capabilities and risks often evolve faster than public understanding.[AI Security Institute]aisi.gov.ukAI Security InstituteFrontier AI Trends Report by The AI Security Institute (AISI)The UK AI Security Institute (AISI) has conducted evalu…

Competition and interoperability

Competition policy may become increasingly important in AI markets.

If users can switch providers easily, move data between systems and avoid lock-in, infrastructure becomes more contestable and less dependent on a single organisation. OECD competition analyses increasingly focus on these questions.[OECD]oecd.orgai openness 02f73362 enAI opennessAug 14, 2025 — This paper analyses current trends in open-weight foundation models using experimental data, illustrating b…

Public-sector expertise

Governments need enough technical capability to evaluate vendors, negotiate contracts and regulate effectively.

Without that expertise, public institutions risk becoming permanently dependent on private providers for core cognitive functions. UNESCO’s public-sector programmes are partly motivated by this concern.[UNESCO]unesco.orgTrust, capacity and motivation: How public administrations…Mar 20, 2026 — Turning to UNESCO's Artificial Intelligence and Digital Tran…

Universal literacy and training

Broad access means little if people cannot evaluate outputs or recognise errors.

Education systems may increasingly need to teach AI literacy alongside traditional digital literacy. UNESCO and other international organisations have repeatedly argued that human judgement and critical evaluation should remain central.[UNESCO]unesco.orggovernments must quickly regulate generative ai schoolsGovernments must quickly regulate Generative AI in schoolsSep 8, 2023 — UNESCO is calling on governments to implement appropriate regulat…[UNESCO]unesdoc.unesco.orgfor generative AI in education and research…

Access beyond affluent regions

If advanced AI becomes concentrated in a small number of countries, global inequality could deepen.

Policies that support infrastructure development, language inclusion, educational access and international research collaboration may therefore become part of the broader effort to make abundant intelligence genuinely global.[UNESCO]unesco.organticipating change how will ai shape future public serviceHow Will AI Shape the Future of Public Service?Dec 19, 2025 — As AI reshapes automation, procurement, and even elements of decision-makin…[OECD.AI]oecd.orgcomponent 6Market features in AI infrastructure: Competition in artificial…Nov 14, 2025 — In the context of AI infrastructure, such competiti…

The deeper question: what kind of civilisation-scale technology is AI?

The fight over broad access ultimately reflects a larger uncertainty about what AI becomes.

If advanced AI remains mainly a productivity tool for corporations, the access question will matter, but within familiar economic frameworks.

If AI develops into something closer to a general-purpose cognitive infrastructure layer for civilisation, the stakes become much larger.

In that world, access to reliable intelligence assistance could influence scientific discovery, educational opportunity, democratic participation, entrepreneurship, public administration and long-run human flourishing. The question would resemble earlier debates about universal schooling, public libraries, electricity grids and internet access: not whether the technology is valuable, but how widely its benefits are distributed.

The strongest version of the AI bloom vision depends on the latter outcome. A future of abundant intelligence is not simply a future with more powerful models. It is a future in which cognitive capability spreads widely enough to expand what ordinary people, institutions and societies can understand, create and achieve.

Whether that happens will depend not only on advances in AI capability, but on ownership, governance, competition, infrastructure, education and public choice. The central political question may be surprisingly simple: when intelligence becomes cheaper, who gets to use it?

Amazon book picks

Further Reading

Books and field guides related to Who Gets Access to Abundant Intelligence?. Use these as the next step if you want deeper reading beyond the article.

Book

The Coming Wave

By Mustafa Suleyman

Directly addresses who controls powerful AI and how access could be governed.

See on Amazon

Book

Power and Progress

By Daron Acemoglu, Simon Johnson

Explores how societies can share gains from transformative technologies.

See on Amazon

Book

Co-Intelligence

By Ethan Mollick

Shows how everyday access to AI can expand practical cognitive capability.

See on Amazon

Book

The Age of A. I.

By Henry Kissinger, Eric Schmidt et al.

Frames AI as a civilisation-scale technology affecting knowledge and power.

See on Amazon

Browse more on Amazon:The Coming WavePower and ProgressCo Intelligence

As an Amazon Associate I earn from qualifying purchases.

eBay marketplace picks

Marketplace Samples

Example marketplace items related to this page. Use the search link to explore similar finds on eBay.

Shop location

UsingUSA USA USAUKCanadaAustraliaIreland

Example eBay listing

TERMINATOR ,SKYNET CYBERDYNE SYSTEMS, SKYNET (Artificial Intelligence) Film Prop

SearcheBay.co.uk: artificial intelligence poster

Browse similar oneBay.co.uk

Example eBay listing

A.I. Artificial Intelligence Movie Film Poster Art Print

SearcheBay.co.uk: artificial intelligence poster

Browse similar oneBay.co.uk

Example eBay listing

A.I. Artificial Intelligence - Jude Law - One Sheet Cinema Poster

SearcheBay.co.uk: artificial intelligence poster

Browse similar oneBay.co.uk

Example eBay listing

A.I. Artificial Intelligence Bluray Steelbook + Original Film Poster

SearcheBay.co.uk: artificial intelligence poster

Browse similar oneBay.co.uk

Browse more oneBay.co.uk

Example items shown for inspiration; availability and pricing can change. Branchoria may earn a commission if you purchase through outbound eBay links.

Endnotes

1. Source: institute.global
Title: sovereignty in the age of ai strategic choices structural dependencies
Link:https://institute.global/insights/tech-and-digitalisation/sovereignty-in-the-age-of-ai-strategic-choices-structural-dependencies

2. Source: oecd.ai
Title: the geopgraphy of ai compute mapping what is available and where
Link:https://oecd.ai/en/wonk/the-geopgraphy-of-ai-compute-mapping-what-is-available-and-where

3. Source: oecd.org
Link:https://www.oecd.org/en/publications/the-impact-of-artificial-intelligence-on-productivity-distribution-and-growth_8d900037-en.html

4. Source: oecd.org
Title: component 6
Link:https://www.oecd.org/en/publications/artificial-intelligence-and-competitive-dynamics-in-downstream-markets_ccf0624a-en/full-report/component-6.html

5. Source: oecd.org
Title: 7376c776 en
Link:https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/06/emerging-divides-in-the-transition-to-artificial-intelligence_eeb5e120/7376c776-en.pdf

6. Source: unesco.org
Title: AI literacy and the new Digital Divide
Link:https://www.unesco.org/ethics-ai/en/articles/ai-literacy-and-new-digital-divide-global-call-action

7. Source: unesco.org
Title: guidance generative ai education and research
Link:https://www.unesco.org/en/articles/guidance-generative-ai-education-and-research

8. Source: arxiv.org
Link:https://arxiv.org/abs/2605.18143

Published: May 18, 2026

9. Source: unesco.org
Link:https://www.unesco.org/en/artificial-intelligence/ai-public-sector

10. Source: unesco.org
Link:https://www.unesco.org/en/articles/trust-capacity-and-motivation-how-public-administrations-are-building-ai-readiness

11. Source: oecd.org
Title: ai openness 02f73362 en
Link:https://www.oecd.org/en/publications/ai-openness_02f73362-en.html

12. Source: oecd.org
Title: component 6
Link:https://www.oecd.org/en/publications/competition-in-artificial-intelligence-infrastructure_623d1874-en/full-report/component-6.html

13. Source: unesco.org
Title: governments must quickly regulate generative ai schools
Link:https://www.unesco.org/en/articles/unesco-governments-must-quickly-regulate-generative-ai-schools

14. Source: oecd.ai
Title: can mid sized economies come together to build frontier ai
Link:https://oecd.ai/en/wonk/can-mid-sized-economies-come-together-to-build-frontier-ai

15. Source: unesdoc.unesco.org
Link:https://unesdoc.unesco.org/ark%3A/48223/pf0000386693

16. Source: unesco.org
Title: anticipating change how will ai shape future public service
Link:https://www.unesco.org/en/articles/anticipating-change-how-will-ai-shape-future-public-service

17. Source: unesco.org
Title: ai and education guidance policy makers
Link:https://www.unesco.org/en/articles/ai-and-education-guidance-policy-makers

18. Source: unesco.org
Link:https://www.unesco.org/en/digital-education

19. Source: unesco.org
Link:https://www.unesco.org/ethics-ai/en/articles/unescos-ai-literacy-training-civil-servants-empowering-ethical-ai-governance-around-world-0

20. Source: unesco.org
Title: Artificial intelligence in education
Link:https://www.unesco.org/en/digital-education/artificial-intelligence

21. Source: oecd.org
Link:https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/06/the-effects-of-generative-ai-on-productivity-innovation-and-entrepreneurship_da1d085d/b21df222-en.pdf

22. Source: oecd.org
Link:https://www.oecd.org/en.html

23. Source: oecd.org
Link:https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/12/ai-and-the-global-productivity-divide_f47026c5/c315ea90-en.pdf

24. Source: oecd.org
Title: component 5
Link:https://www.oecd.org/en/publications/artificial-intelligence-and-competitive-dynamics-in-downstream-markets_ccf0624a-en/full-report/component-5.html

25. Source: oecd.org
Link:https://www.oecd.org/en/publications/the-effects-of-generative-ai-on-productivity-innovation-and-entrepreneurship_b21df222-en.html

26. Source: oecd.org
Link:https://www.oecd.org/en/topics/sub-issues/ai-compute.html

27. Source: oecd.org
Title: unlocking productivity with generative ai evidence from experimental studies
Link:https://www.oecd.org/en/blogs/2025/07/unlocking-productivity-with-generative-ai-evidence-from-experimental-studies.html

28. Source: wp.oecd.ai
Title: 20250128 GPAI GenAI FoW report final VOECD
Link:https://wp.oecd.ai/app/uploads/2025/01/20250128_GPAI_GenAI_FoW_report_final_VOECD.pdf

29. Source: arxiv.org
Link:https://arxiv.org/html/2602.10900v3

30. Source: policylabs.frontiersin.org
Link:https://policylabs.frontiersin.org/content/policy-outlook-designing-capability-institutions-and-the-future-of-emerging-compute

Published: December 16, 2025

31. Source: businessinsider.com
Link:https://www.businessinsider.com/ai-fastest-tech-in-history-microsoft-warns-billions-left-out

32. Source: aisi.gov.uk
Link:https://www.aisi.gov.uk/frontier-ai-trends-report

33. Source: Wikipedia
Link:https://en.wikipedia.org/wiki/OECD

34. Source: unesco.org.uk
Link:https://unesco.org.uk/resources/guidance-for-generative-ai-in-education-and-research

35. Source: researchgate.net
Link:https://www.researchgate.net/publication/385105010_UNESCO_Proposal_for_the_use_of_Generative_AI_in_Education_Eight_Challenges_and_Seven_Actions

36. Source: govtech.com
Title: unesco reassert public control over generative ai
Link:https://www.govtech.com/education/k-12/unesco-reassert-public-control-over-generative-ai

37. Source: ohchr.org
Link:https://www.ohchr.org/sites/default/files/documents/issues/education/cfis/ai-education/submissions/subm-artificial-intelligence-un-enti-unesco-ibe.docx

38. Source: verifywise.ai
Link:https://verifywise.ai/ai-governance-library/sector-specific-governance/unesco-ai-education

39. Source: youtube.com
Link:https://www.youtube.com/oecd

40. Source: facebook.com
Link:https://www.facebook.com/theOECD/

Additional References

41. Source: researchgate.net
Link:https://www.researchgate.net/publication/399504193_AI_AS_PUBLIC_INFRASTRUCTURE_A_CRITICAL_REVIEW_OF_THE_TRANSITION_FROM_TOOL_TO_SOCIETAL_NECESSITY

42. Source: brookings.edu
Link:https://www.brookings.edu/articles/ai-growth-acceleration-versus-distributional-fairness/

43. Source: discovery.ucl.ac.uk
Title: Miao and Holmes 2023 Guidance for Generative AI in Education and Resear
Link:https://discovery.ucl.ac.uk/id/eprint/10176438/1/Miao%20and%20Holmes%20-%202023%20-%20Guidance%20for%20Generative%20AI%20in%20Education%20and%20Resear.pdf

44. Source: laweconcenter.org
Title: ai productivity and labor markets a review of the empirical evidence
Link:https://laweconcenter.org/resources/ai-productivity-and-labor-markets-a-review-of-the-empirical-evidence/

45. Source: linkedin.com
Title: [ai wealth]({{ ‘ai-wealth/’ | relative_url }}) power polarization digital divides new andre iedge
Link:https://www.linkedin.com/pulse/ai-wealth-power-polarization-digital-divides-new-andre-iedge

46. Source: weforum.org
Title: World Economic Forum Generative AI has disrupted education
Link:https://www.weforum.org/stories/2023/09/generative-ai-education-unesco/

47. Source: aigi.ox.ac.uk
Title: a blueprint for multinational advanced ai development
Link:https://aigi.ox.ac.uk/publications/a-blueprint-for-multinational-advanced-ai-development/

48. Source: theschoolhouse.org
Title: Guidance for Generative AI in Education and Research
Link:https://www.theschoolhouse.org/post/generative-ai-in-education-research-guidance

Published: May 12, 2025

49. Source: GOV.UK
Title: frontier ai capabilities and risks discussion paper
Link:https://www.gov.uk/government/publications/frontier-ai-capabilities-and-risks-discussion-paper/frontier-ai-capabilities-and-risks-discussion-paper

50. Source: school-education.ec.europa.eu
Title: guidance generative ai education and research
Link:https://school-education.ec.europa.eu/en/discover/publications/guidance-generative-ai-education-and-research