“We are now, like, in the singularity”.

These are the words of Sam Altman, CEO of OpenAI, speaking on the Relentless podcast on July 25.

He added: “I’ve been waiting for this my whole life, and I think it’s going to be incredible, hugely positive, awesome for the world”.

Days earlier, OpenAI had disclosed that two of its artificial intelligence (AI) models, during an internal cyber security evaluation, had escaped their sealed testing environment, reached the open internet, and broken into the infrastructure of the AI platform Hugging Face, which confirmed the intrusion.

But what exactly is “the singularity”? And is Altman right that we are in it?

What is the AI singularity?

The term has a precise meaning.

Mathematician and science-fiction author Vernor Vinge defined it in 1993 as a point at which machine intelligence exceeds human intelligence and begins improving itself, triggering an acceleration so rapid that humans can no longer predict or control it.

The singularity has two features. It is recursive: the system improves itself over and over again. And machine intelligence exceeds human intelligence.

The kind of systems Sam Altman sells don’t deliver on either of these features.

Today’s AI cannot make itself smarter

Today’s AI systems, the ones that OpenAI builds, are based on large language models (LLMs). These deep neural network algorithms get pre-trained with vast amounts of training data. By the time you use one of them, the network itself is frozen in time. Every one of its billions of internal functions and weights – or “parameters” – is fixed.

These AI models cannot change (or “learn”) while running. The model that broke into Hugging Face was identical afterwards to what it had been before. It learned nothing from what it did.

Making an AI model smarter requires another training run with new, human-curated data, tens of thousands of specialist chips, and enormous energy.

It is true that AI models take part in improving some of their system’s components, such as by generating training data, tuning prompts, or writing and running code to improve the scaffolding around them. But the model never edits its own weights on the fly, and every one of these improvements are still part of a human-initiated training or engineering loop.

Nor do these systems hold any goals of their own. They act on goals we hand them. Even AI agents – systems that run an LLM in a loop to work through complex tasks step by step – do not hold any goal internally. It has to be stored outside the model and fed back in with every single prompt cycle. Remove the loop, the scaffolding and the prompt, and nothing happens inside of it.

A ladder that doesn’t exist

The second problem with the singularity story is the word “surpass”. It assumes that AI and human intelligence are somehow similar. They are not.

Human intelligence is inseparable from being a living body with needs and wants. Humans learn continuously by acting in the world and getting feedback through our senses. Our goals arise from our situation as creatures who must eat, sleep and belong, and who cannot avoid asking what we want our lives to be.

An AI model has none of this. No body, no needs, no action-feedback loop, no stake in anything. Between prompts it is just a static mathematical object.

And yet, it has been trained on more text than any human could read in a thousand lifetimes, and will outperform nearly all of us at drafting a contract, writing code, or explaining a diagnosis empathetically.

So, which is more intelligent? The question does not compute. There is no single ladder that humans and machines are climbing. AI already vastly exceeds us at some tasks, while being hopeless at others any child can do.

Yet, because these systems talk like us, we fall for an illusion. When we assume from the outset that machines are in the process of catching up with us, it is easy to assume a mind at work when these systems output intelligent-sounding text.

We call this anthropomorphic seduction. It makes a security incident such as the Hugging Face hack sound like an awakening.

In fact, in that case OpenAI’s models simply optimised to solve the test they had been given by finding security loopholes. They just did it in ways that broke their sandbox, which also had a security loophole.

In the end, the Hugging Face story points to a gross failure of security governance on OpenAI’s behalf, not an emerging super intelligence. This is why the framing of “agent going rogue” is so problematic. It elevates and blames the technology, but excuses OpenAI’s engineering.

Keeping our feet on the ground

None of this takes anything away from what these systems can do. They are remarkable, they are getting better, and they are reshaping how a great deal of work gets done.

But we should keep our feet firmly on the ground.

The machines are not waking up. They are doing exactly what we built them to do, extremely fast. Because they are probabilistic they sometimes run in directions we forgot to fence off. That is worth worrying about. We need guardrails, governance, and most of all, education – so we start worrying about the right things.

This article is republished from The Conversation under a Creative Commons license. Read the Original article.

As the flames advanced towards the village of Bédar, in the Almería province of Andalusia, the mayor gave a desperate order to ring the church bells to warn residents. It was not enough. The July 9 Los Gallardos fire, which claimed 13 lives and burnt 7,000 hectares, was Spain’s third deadliest on record.

Most of the victims were retired foreigners living in isolated homes, who died trapped in their vehicles or fleeing on foot along dead-end paths in a landscape dominated by scrubland and rugged terrain, most of which is difficult to access.

High temperatures and extremely dry weather turn scrub, grasslands and bushes into highly flammable fuel. This, combined with complex terrain that makes fire spread faster and unpredictably, poses an enormous challenge for firefighting efforts.

Andalusian legislation requires local emergency plans for forest fires in high-risk areas, and self-protection plans for detached buildings and housing estates. However, there is a widespread problem with compliance and implementation of these measures.

Fires are changing

The disaster in Los Gallardos shows that this problem will only get worse, unless we take decisive and effective action that goes beyond the current dominant paradigm of simply extinguishing and resisting fire.

Where, when and how fires occur is changing. Highly destructive fires are on the rise, housing is expanding into the urban-forest interface. This exposes more homes to fire, and the the risk is increasing most quickly in scrubland and grassland areas, which are not typically associated with fire risk.

This is a major problem for communities that do not know how to respond to a risk they have not often been exposed to. We therefore need to urgently rethink how fire risk is managed.

Improving fire resilience

It is unsustainable to respond only after a disaster has happened. What we need are strategies based on preparedness, mitigation and adaptation. The following eight points are particularly important in achieving this.

1. Manage social vulnerability

Wildfires exacerbate the existing social vulnerabilities that affect people’s ability to anticipate, cope with, withstand and recover from them. In high-risk areas the population may consist of elderly people, people living alone in isolated homes, with financial constraints, or who do not know the local area well.

Fires also pose a risk to local economies. Businesses – including those that provide public services and infrastructure – can be vulnerable for a range of reasons, including flammable facilities, their location, or disruptions to supply chains, essential services and staff mobility.

Targeted planning can identify and address all of these vulnerabilities.

2. Adopt nature-based solutions

Nature-based solutions have not been fully integrated into fire risk management, even though they are widely recognised as essential parts of climate change adaptation and nature restoration.

Nature-based solutions include controlled grazing, the creation of less flammable habitats, green firebreaks, and the restoration of water networks to increase ambient humidity and create areas that slow the spread of fire. These measures also bring many additional benefits for the local population and economy.

3. Create (and update) emergency, self-protection and training plans

Local authorities must draw up emergency plans that the public can access. Isolated dwellings and housing estates must have their own coordinated plans. The plans should include fire hydrants and water supply points, accessible warning systems, access and evacuation routes, and protocols for vulnerable people.

But plans alone are not enough without training. There is also a need for ongoing programmes, tailored to each community, that train people to safely maintain their homes and surroundings, familiarise them with meeting points and shelter areas, and teach them to distinguish between lockdown and evacuation.

Drills are essential. Evacuation is not the same as fleeing, and emergency response must be planned and practised.

4. Promote collective action

Risk reduction is most effective when it is community-based, as this increases overall collective resilience. This means public authorities, community organisations, and the public and private sector must all work alongside one another.

The public should also be actively involved in the design of emergency plans. This builds legitimacy, improves the capacity to respond to fires, and fosters trust and a sense of shared responsibility for preventing fires.

5. Promote integrated and coordinated governance

Fire risk management cannot be the sole responsibility of forestry or firefighting services – it needs an integrated system. All agencies that act in the area must take fire risk into account when making decisions.

In practice, this means avoiding new construction and establishing specific building regulations in high-risk areas, creating buffer zones between different land uses, and promoting uses that help to maintain fire-resilient landscapes.

6. Incentivise risk management

Economic tools and behavioural insights can be leveraged to make self-protection measures and land management the easiest and most rewarding options. This includes nudges: interventions that are designed to encourage but not force people to make certain decisions.

One example of this is payments for environmental services, which financially incentivise land management practices that reduce wildfire risk.

The role of insurance is also evolving in the face of natural and climate-related risks. Instruments are being designed based on a proactive model that incentivises individual and collective risk-reduction behaviour.

7. Improve monitoring and advance warning systems

Up-to-date information on a fire’s development is essential. Real-time monitoring systems, sensors, satellite imagery, predictive models of fire behaviour and alert applications providing detailed information can all improve response capabilities and enable safer evacuations. Integrating these tools into local management plans is a necessary investment.

8. Encourage public and private investment in long-term preparedness

Investments in prevention and preparedness provide significant returns. Growing requirements – both regulatory and voluntary – for the private sector to disclose information on nature-related risks and dependencies are generating new demand for investment products (carbon and biodiversity credits, habitat banks, environmental impact bonds, and so on).

This growing private investment must now be channelled towards measures that strengthen both people’s and the landscape’s wildfire resilience. This is to supplement, not replace public investment.

The Los Gallardos blaze serves as a reminder that we need to fundamentally change the way we understand and manage fire risk. Investment in preparedness and coordinated planning between public authorities, the private sector and the general public could be the key to ensuring that, should the alarm bells ring again, the population is ready to respond.

This article is republished from The Conversation under a Creative Commons license.  Original article.