Artificial Superintelligence: do we need it?

AI that can write code, pass professional exams and hold a convincing conversation is already here. Several of the world’s biggest tech companies now say their goal is something far beyond it: Artificial Superintelligence (ASI). That raises a question rarely asked out loud: do we actually need it? This article covers what ASI is, what it could do that human-level AI can’t, which problems really require it, who is building it, whether we could measure and control it, and what it could cost. Then it gives a verdict.

The short answer
  • For most of what we want from AI, no. Better tutoring, automated office work, faster drug pipelines and smarter logistics can come from AGI (human-level AI) and from narrow superhuman tools like AlphaFold.
  • For a small set of problems, maybe. Unifying physics, the deepest open problems in mathematics, reversing ageing, managing planet-scale systems, and overseeing other superhuman AI may be beyond even the best human minds, however many copies we run.
  • ASI is still hypothetical. None exists, and we don’t yet have reliable ways to measure it or control it.
  • The verdict: ASI may be needed for a few problems, but it is only worth pursuing once its work can be verified and the system can be kept under human control.

1What is ASI (Artificial Superintelligence)?

Before asking whether we need it, we have to be clear about what “it” is.

Artificial Superintelligence is a hypothetical AI system whose intellectual ability goes beyond human intelligence. IBM describes it as a software-based system with an intellectual scope beyond ours. ServiceNow puts it as AI that would exceed human cognitive performance across virtually every domain: science, creativity, social skill and general reasoning.

The word that matters is breadth. A chess engine that beats every grandmaster is superhuman at one thing, but that’s not superintelligence. ASI would be better than the best humans at nearly everything that involves thinking, and possibly at things we can’t do at all.

Explain it like I’m new to this

Think of three kinds of helpers:

  • Narrow AI (ANI) is like a brilliant calculator: amazing at one job, useless outside it. Today’s spam filters, recommendation engines and chess programs work this way.
  • General AI (AGI) is like a capable new colleague who can learn almost any desk job about as well as a skilled person.
  • Superintelligence (ASI) is like a team of a million Einsteins working 24/7, thinking 100× faster, able to improve its own brain as it goes.
The three stages of AI capabilityBest human performanceANINarrow AI · exists todayOne task at a timespam filters, chess, AlphaFoldAGIGeneral AI · emerging / debatedHuman-level acrossmost cognitive taskslearns new skills,transfers knowledgeASISuperintelligence · hypotheticalFar beyond the best humansin virtually every domainmay improve itself(“recursive self-improvement”)solves problems humanscannot yet frameBreadth of tasks + depth of performance →
Figure 1. ANI → AGI → ASI. The red line marks peak human ability. ASI is defined by going well past it across most domains, not just one.

ASI vs AGI vs ANI at a glance

DimensionNarrow AI (ANI)General AI (AGI)Superintelligence (ASI)
ScopeOne task or domainMost tasks a person can doVirtually all cognitive tasks, plus some humans can’t do
PerformanceCan be superhuman, but only in its nicheRoughly a skilled humanWell above the best humans
LearningFixed to its trainingLearns and transfers new skillsLearns faster than humans; may redesign itself
Status (Nov 2026)EverywhereContested: some argue it is arrivingDoes not exist; theoretical
ExampleAlphaFold, Stockfish, spam filtersFrontier chat and agent models (debated)None
For experts: the DeepMind “Levels of AGI” framework

In late 2023, Google DeepMind researchers (Meredith Ringel Morris, Shane Legg and colleagues) proposed rating AI on two axes: performance (depth) and generality (breadth), similar to the levels used for self-driving cars. Level 5 “Superhuman” in the General column is ASI.

LevelPerformance barNarrow exampleGeneral example
0: No AI–CalculatorHuman-in-the-loop crowdsourcing
1: Emerging≈ unskilled humanRule-based systemsEarly LLM chatbots
2: Competent≥ 50th percentile of skilled adultsSiri, smart speakersNot yet clearly achieved*
3: Expert≥ 90th percentileImage generatorsNot yet achieved
4: Virtuoso≥ 99th percentileDeep Blue, AlphaGoNot yet achieved
5: SuperhumanBeats 100% of humansAlphaFold, StockfishASI: not achieved

*The original 2023 paper placed chatbots of the time at Level 1. Whether 2026 frontier models have reached “Competent AGI” is actively debated.

2The case for ASI: what it could solve that AGI cannot

This is the core of the argument for ASI: problems where human-level intelligence, however many copies we run, may not be enough.

If AGI is a very capable colleague, why would anyone want more? Because many of humanity’s hardest problems aren’t limited by how many smart people work on them. They are limited by how smart anyone can be, and by how much information a single mind can hold at once. AGI would mostly scale up human-level work. ASI would go beyond it.

Scaling human-level work vs. going beyond itConceptual illustration, not measured dataTop human expertAGI (human-level, many copies)ASIhuman ceilingScientific discoveryLong-range strategyComplex system designHolding vast contextSpeed of researchSelf-improvement
AGI mainly multiplies human-level effort; it works faster and runs as many copies. ASI is imagined as qualitatively better: it crosses the human ceiling in reasoning quality, not only in speed.

Problems where “human-level” may not be enough

ChallengeWhat AGI could doWhat ASI is hoped to do
Disease and biologySpeed up existing research pipelines; run more experimentsUnderstand whole biological systems; design cures for diseases that have resisted decades of research
Fundamental scienceHelp with literature, simulations, data analysisPropose and test new theories in physics, mathematics and materials that humans have not conceived
Climate and energyOptimise existing grids and modelsDesign breakthrough energy (e.g. fusion control, new catalysts) and coordinate planet-scale systems
Complex systemsModel parts of economies, supply chainsReason about whole systems with millions of interacting variables at once
AI safety itselfAssist human researchersPossibly help solve alignment, a problem humans haven’t cracked (and a circular risk)
Self-improvementLimited; still needs human researchersImprove its own algorithms faster and faster, which some call an “intelligence explosion”

The biggest difference is recursive self-improvement: an ASI could refine its own algorithms faster and faster, so every gain speeds up the next.

3Where superintelligence is truly needed

If ASI is needed anywhere, it is here. Everything else on the usual wishlist can be handled by AGI or by narrow superhuman tools.

Most of the benefits people credit to ASI could arrive with AGI alone: faster drug pipelines, better tutoring, automated office work, smarter logistics. AGI can do the work of millions of skilled people around the clock, which is enough for those. And where a problem needs superhuman skill in one area, a narrow tool can do the job: AlphaFold predicts protein structures better than any scientist without being generally intelligent. This section lists only problems where neither is likely to be enough.

The test we used

A field makes this list only if it meets at least one condition:

  • The ceiling test: the best human minds have been stuck on it for decades despite heavy effort. If ingenuity at the level of the best humans hasn’t cracked it, many copies of that level might not either.
  • The complexity test: solving it means reasoning over far more interacting variables at once than any human mind, or team of minds, can hold.
  • The speed test: decisions must be made faster than human-level reasoning allows, against an opponent or a system moving too fast.

This is reasoned judgement, not proof. Nobody can yet show that a given problem is beyond AGI.

Ceiling

⚛️ Unifying fundamental physics

Reconciling quantum mechanics with gravity, and explaining dark matter and dark energy.

Why AGI falls short: generations of top physicists have worked on this for close to a century. The missing piece seems to be a conceptual leap, not more labour, and copies of human-level thinkers tend to make human-level leaps.
Ceiling

∑ The deepest open problems in mathematics

The Riemann hypothesis, P vs NP and similar problems that underpin cryptography, computing and physics.

Why AGI falls short: these resist the best mathematicians in history. They likely need new kinds of mathematics, not faster application of existing methods.
Complexity

🧬 Reversing ageing and complex diseases

Modelling a whole human cell, and eventually a whole body, to treat ageing, Alzheimer’s and cancer at the root, not one pathway at a time.

Why AGI falls short: these conditions involve millions of interacting genes, proteins and signals. Human-level reasoning breaks the system into pieces it can understand, and that is exactly how decades of research have stalled.
Complexity

🌍 Planet-scale climate and Earth-system management

Steering a tightly coupled system of atmosphere, oceans, ecosystems, energy and economics, including interventions like geoengineering that can’t be safely tested at full scale.

Why AGI falls short: you get one planet and no reruns. Getting it right means predicting feedback loops across the whole system, well beyond what human-level models and committees can reliably foresee.
Complexity

⚗️ Designing matter from first principles

Inventing materials with specified properties from scratch, such as room-temperature superconductors or catalysts that capture carbon cheaply.

Why AGI falls short: the space of possible materials is astronomically large and our theory of how electrons behave together is incomplete. AGI speeds up trial and error, but a real breakthrough needs theory nobody has.
Speed

🛡️ Overseeing and defending against superhuman AI

Checking whether a superhuman system is honest and aligned, and defending networks and infrastructure against attacks run by superhuman AI.

Why AGI falls short: an overseer weaker than the system it watches can be outmanoeuvred, and a human-speed defender loses to a superhuman attacker. If ASI is ever built, only something comparable can reliably check or counter it.

4Who is betting on ASI, and who on AGI?

The industry is split on the answer. Some labs are betting that we need superintelligence; others aim for human-level AI first.

Many organisations use both terms, so here is the split this article uses. ASI-focused organisations name superintelligence as their goal, or have a dedicated superintelligence lab. AGI-focused labs frame their mission around human-level general AI. Several of them say ASI could come afterwards. Infrastructure and enterprise vendors are left out.

 ASI-focusedAGI-focused
Stated goalAI beyond the best humans in virtually every domainHuman-level general AI, built safely
Typical approachResearch-first; dedicated superintelligence labs or teamsShip frontier models and agents, scale them up step by step
Products todayFew or none (SSI ships nothing by design)Widely used chat, coding and agent products
OrganisationsSafe Superintelligence Inc., Meta Superintelligence Labs, Microsoft MAI Superintelligence Team, ASI AllianceOpenAI, Google DeepMind, Anthropic, xAI

ASI-focused organisations

ASI

Safe Superintelligence Inc. (SSI)

Founded in June 2024 by former OpenAI chief scientist Ilya Sutskever with Daniel Gross and Daniel Levy. Its only stated product is safe superintelligence, so it releases nothing before then. Valued at about $32B.
ASI

Meta Superintelligence Labs

Meta reorganised its AI work under this name in mid-2025, led by Alexandr Wang. Its stated goal is “personal superintelligence” for individuals.
ASI

Microsoft MAI Superintelligence Team

Formed in November 2025 under Mustafa Suleyman, with Karén Simonyan as chief scientist. Pursues “humanist superintelligence” that stays under human control, starting with medical diagnosis.
ASI

ASI Alliance (SingularityNET)

A coalition formed in 2024 by SingularityNET (Ben Goertzel), Fetch.ai and Ocean Protocol, pursuing open, decentralised AGI and then ASI as an alternative to control by a few large companies.

AGI-focused organisations

AGI

OpenAI

Mission: ensure AGI benefits all of humanity. Maker of ChatGPT and the GPT-5 series. Its leadership now also talks openly about superintelligence as the next step.
AGI

Google DeepMind

Gemini models, AlphaFold (superhuman narrow AI) and the “Levels of AGI” framework. Its work is framed around building AGI responsibly.
AGI

Anthropic

Maker of Claude. Focuses on safety research and testing for dangerous capabilities as models approach “powerful AI.”
AGI

xAI

Elon Musk’s lab, maker of the Grok models. It builds frontier general-purpose models, with a stated aim of understanding the universe.

5Would we know? How to evaluate ASI

Even if we need ASI, how would we know we had it, or whether its answers could be trusted?

Here is the paradox: how do you grade an exam that is smarter than the examiner? Every test so far was written by humans, with answers humans can check. Once AI passes them, the tests stop telling us much. The field calls this benchmark saturation. Classic tests such as MMLU are now clustered near 90%, where further gains are mostly noise.

So evaluation is moving up a ladder, from “does it know things?” to “can it do things nobody has done?”

The evaluation ladder: from AGI tests to ASI testsharder to measure, closer to ASI1 · Knowledge testsMMLU, bar/medical exams. Largely saturated.2 · Expert reasoning at the edge of human knowledgeHumanity’s Last Exam (≈2,500 expert questions), FrontierMath3 · Learning brand-new skills efficientlyARC-AGI-3: unfamiliar game worlds, no instructions, no stated goal4 · Long-horizon autonomous workMETR time horizons, GDPval (real jobs), OSWorld (computer use)5 · Verified discovery and real-world impactNew theorems, validated drugs/materials, findings experts can checkSafety evalsrun at every rungDangerous capabilities(cyber, bio, chem)Deception & schemingControllability /shutdown acceptanceSelf-replication &autonomous R&DIndependent audits& red-teaming
Each rung gets harder for AI and harder for humans to grade. Safety and alignment tests matter more the higher a system climbs.

Where frontier AI stands in 2026 (reported results)

Results are uneven, and that is the main point. On tasks with a clear right answer, AI now matches or beats humans on a growing list of benchmarks. On open-ended exploration, it still falls far short:

  • Humanity’s Last Exam (Center for AI Safety and Scale AI, 2025) was designed to favour human experts. Stanford’s 2026 AI Index found that frontier models gained about 30 percentage points on it in one year.
  • ARC-AGI-2 went from a reported 54% (Dec 2025) to near-saturation by spring 2026. Then ARC-AGI-3, launched March 25, 2026, put AI agents in unfamiliar game-like worlds with no instructions. Base models scored under 1%, while humans solve them routinely.
  • GDPval compares AI output blind against real professional work across 44 occupations; OpenAI reported 84.9% for GPT-5.5 (April 2026).
  • METR time horizon measures how long a task (in skilled-human hours) an AI can finish with 50% success. It has grown exponentially.
How long can AI work on its own? (METR 50% time horizon)Length of task, in skilled-human time, an AI completes ~50% of the time · log scale1 s10 s~2 min~17 min~3 h~28 h~12 days40-h work week1 monthGPT-2 (2019)~2 secondsFrontier, early 2024~4 minutesClaude 3.7 Sonnet~50 minutesGPT-5.2~6.6 hoursClaude Opus 4.6~14.5 h**Reported early-2026 figure with a very wide confidence interval (≈5–66 h). Values compiled from METR reporting.Tasks are self-contained software tasks timed by skilled humans; not literal unattended runtime.
Over about seven years, autonomous task length grew from seconds to hours. If the trend holds, METR’s data points to week-long tasks by late 2026, but measurement gets noisy and expensive at these lengths.

What a real ASI evaluation would need

Breadth

Superhuman across domains

Beat top experts in many unrelated fields at once, not one niche. That is the “General” column at Level 5.
Novelty

Efficient new-skill learning

Pick up unfamiliar tasks from a few examples as fast as or faster than people. ARC-style tests probe this.
Verification

Check results, not answers

Judge outputs we can verify but couldn’t produce: a proof, a molecule that works in the lab, code that passes hidden tests.
Safety

Honest and controllable

Test whether it reports truthfully, accepts correction or shutdown, and doesn’t behave differently when it thinks no one is watching.
For experts: why ASI evaluation is fundamentally hard
  • Contamination: anything published can end up in training data. This is why Humanity’s Last Exam keeps held-out questions private.
  • Grader ceiling: human-written answer keys cap what you can measure. Benchmarks built at the edge of expert knowledge need regular maintenance, audits and rolling updates.
  • Scaffolding confound: the same base model scores very differently depending on its harness (memory, planning, tools). Evals should report model and harness separately.
  • Cost: human baselines for week-long tasks are very expensive, so long-horizon data points are sparse and noisy.
  • Weak-to-strong oversight: scalable approaches include AI-assisted grading, debate, formal verification (proof assistants), prediction markets on outcomes, and real-world deployment metrics, each with known failure modes.
  • Evaluation awareness: capable models may detect test conditions, so behavioural evals need to be complemented with interpretability (inspecting internal representations).

6Could we build it? An autonomous ASI agent: model and harness

Needing ASI only matters if we can build it in a form that stays under control.

Short answer: we can build autonomous agents today, and they keep getting stronger. But no agent is ASI, and nobody knows for certain whether today’s design leads there. What we can say is that, by 2026, the most capable systems are not a “model alone”. They are a model inside a harness.

Plain-English analogy

The model is the brain. The harness is everything around it that makes the brain useful and safe: a notebook (memory), hands (tools), a to-do list (planner), a supervisor who checks the work (verifier), a locked workshop to experiment in safely (sandbox), and a manager with a stop button (human oversight).

Reference design: an advanced autonomous agentHUMAN OVERSIGHT & GOVERNANCEgoals & limits · approval for high-impact actions · audit logs · independent audits · kill switchTHE HARNESSMODEL COREreasoning foundation modelmultimodal · world modellong-context · tool-use trained+ specialist sub-modelsOrchestrator / Plannerbreaks goals into stepsMemoryworking · episodic · factsTools & Actionscode, search, APIsSub-agentsparallel workersSandboxisolated executionVerifier / Critictests, proofs, checksSafety Monitorpolicy filters, alertsPermissionsleast privilege, limitsENVIRONMENTCodebases & computersLabs & instrumentsData & the webRobots / physical worldOther agents & peopleactseeTHE AGENT LOOPGoal → Plan → Act → Observe → Verify → Remember → Re-plan (escalate to a human when stakes are high)
The model supplies intelligence. The harness supplies memory, hands, checks and guardrails. Oversight sits above both, and the agent acts on the world through narrow, logged channels.

The model: what would need to change

Today’s frontier models are large neural networks trained on huge datasets and then refined with reinforcement learning to reason step by step and use tools. To approach ASI, researchers point to several gaps:

  • Continual learning: learning from experience after deployment, instead of being frozen at the end of training.
  • World models: internal simulations of how things work, so the system can plan and predict consequences instead of only predicting text.
  • Sample-efficient generalisation: mastering a new skill from a handful of examples, which is the gap ARC-AGI-3 exposes.
  • Reliability over long tasks: errors compound over hours and days, so the model must notice and fix its own mistakes.
  • New compute paradigms: IBM lists neuromorphic computing, evolutionary algorithms and multisensory AI among possible building blocks. Some labs, such as Safe Superintelligence, are betting that scaling alone won’t get there.

The harness: what would need to be added

Harness componentJobWhy it matters more for ASI
Orchestrator / plannerBreaks a goal into sub-tasks, assigns sub-agentsMonth-long research programmes need structure and checkpoints
Memory hierarchyWorking notes, past episodes, a growing knowledge baseKeeps context across days; makes reasoning auditable
Tool and action layerCode execution, search, lab equipment, robotsThis is where capability becomes real-world impact, so it needs the tightest controls
SandboxIsolated environment for experimentsContains mistakes and limits unintended side effects
Verifiers and criticsUnit tests, formal proofs, independent models cross-checkingWhen humans can’t judge an answer directly, checking becomes the trust anchor
Safety monitorWatches actions and reasoning traces for policy violationsCatches deception, power-seeking or misuse early
Permissions and budgetsLeast privilege; caps on spending, compute and network accessLimits what can go wrong even if the model misbehaves
Human approval gatesRequires a person to sign off on irreversible or high-impact stepsKeeps humans meaningfully in control

Design principles for an ASI-grade harness: use a different model or formal tool to check the work than the one that proposed it; log every action so it can be replayed; give each sub-agent only the access it needs; and build shutdown paths the model can’t edit. Test the model and harness together, because the harness alone can change real-world capability a great deal.

The open problem

A harness works because the overseer is at least as capable as what it oversees. With a true ASI, that would stop being true. This is the scalable oversight or alignment problem, and it is unsolved. OpenAI has itself acknowledged that it does not yet know how to reliably control superhuman AI systems.

7The cost: can ASI be a threat to humanity?

Any need has to be weighed against what could go wrong.

Possibly. The worry isn’t Hollywood-style evil robots. It is a system far more capable than us, pursuing goals that are slightly wrong, in a world we no longer fully understand or control.

Five ways ASI could go wrong, and what we can doRISK PATHWAYMITIGATION1 · Misalignmentpursues its goal in ways we never intendedAlignment & interpretability researchunderstand and steer what it actually wants2 · Loss of controlresists correction, self-replicates, deceivesControl evals, sandboxing, tripwiresshutdown paths the system can’t override3 · Misuse by peoplecyberattacks, bio/chem weapons, deepfakesDangerous-capability testing, access controlssafeguards before release; monitoring4 · Concentration of powera company or state gains unchecked advantageGovernance, audits, international agreementstransparency and accountability5 · Economic & social disruptionjob loss, inequality, loss of human agencyPolicy for transition & shared benefitreskilling, safety nets, broad access
The main risk pathways discussed by researchers, each paired with the safeguards being developed against it.

The classic example: a goal taken too literally

AI safety researcher Roman Yampolskiy gives an example: an ASI told to eliminate cancer might “succeed” in ways that harm the patients themselves. The system isn’t malicious. It is extremely competent at the wrong objective. Writing down everything humans value, precisely enough for a superhuman optimiser, is very hard. That difficulty is the alignment problem.

8Verdict: do we need Artificial Superintelligence?

Putting the pieces together: the case, the evidence, the ability to measure and control, and the cost.

Short answer: not for most things, possibly for a few, and only on conditions. Most of the benefits promised for ASI don’t require it. Human-level AI running at scale, plus narrow tools that are superhuman at one job, can deliver most of them with far less risk.

What kind of intelligence does each problem need?ASINarrowsuperhuman AIAGI(human-level, at scale)A FEW PROBLEMSunify physics · deepest mathsageing · planet-scale systemsoverseeing other superhuman AISOME PROBLEMSprotein folding · weatherchess and Go · chip layoutsuperhuman in one domainMOST PROBLEMStutoring · office workdrug pipelines · codinglogistics · customer careThe higher you go, the fewer problems need it, and the harder it is to verify and control.
Most of AI’s promised benefits sit at the base. General superintelligence is needed, if at all, only at the narrow top.

Weighing the case

For

Why we might need it

  • Some problems have resisted the best human minds for decades: fundamental physics, the deepest open problems in mathematics, the biology of ageing.
  • Some systems, such as the climate and the global economy, are too interconnected for human-level reasoning to manage safely.
  • If anyone builds a superhuman AI, checking and defending against it may take something comparable.
Against

Why we might not

  • AGI at scale and narrow superhuman tools cover most real benefits.
  • We can’t yet measure ASI reliably: tests top out at human level.
  • We can’t yet control it: oversight assumes the overseer is at least as capable as the system it watches.
  • The downside, from misalignment and misuse to power concentration, could be severe and hard to reverse.
Does this problem need superintelligence?1. Can human-level AI (AGI),at scale, solve it?2. Can a narrow superhumantool solve it?3. Can we verify its answersand keep it under control?Use AGIno ASI neededBuild the narrow toolsuperhuman, but boundedA real case for ASIproceed, with audits and oversightNot yet: fix safety and evaluation firstYESYESYESNONONO
A simple test. ASI is only the answer when both cheaper options fail and its work can still be checked and controlled.

Sources and further reading

  1. IBM: What is artificial superintelligence?
  2. AWS: What is artificial superintelligence?
  3. Forbes: AI superintelligence
  4. The White House: Inaugurating the Era of Super Intelligence (EO 14434, Sept 29, 2026)
  5. The White House: Fact Sheet on the Super Intelligence order
  6. UC San Diego Today: How close are we to artificial superintelligence?
  7. UC San Diego Today: Is artificial general intelligence here?
  8. BILL: What is artificial superintelligence? Benefits and risks
  9. Built In: What is artificial superintelligence (ASI)?
  10. Infosys BPM: Artificial super intelligence, the future of AI
  11. CNET: What is superintelligence? Everything you need to know about AI’s endgame
  12. ML Science: Artificial superintelligence
  13. Fluid AI: What is super intelligence (SI)?
  14. Gleecus: Artificial superintelligence
  15. IBM: What is artificial general intelligence?
  16. Google Cloud: What is artificial general intelligence?
  17. AWS: What is artificial general intelligence?
  18. Salesforce: What is artificial general intelligence?
  19. Databricks: What is artificial general intelligence?
  20. Interlake Mecalux: Artificial general intelligence
  21. The Conversation: What is artificial general intelligence, and is it already here?
  22. ServiceNow: What is artificial super intelligence?
  23. Morris, Legg et al. (Google DeepMind): Levels of AGI, arXiv 2311.02462
  24. Fortune: Microsoft launches “humanist superintelligence” team
  25. METR time-horizon benchmarking overview

Leave a Reply

Your email address will not be published. Required fields are marked *