Une intelligence artificielle pourrait-elle provoquer la disparition de l’humanité ? Superintelligence, armes biologiques, guerre nucléaire… cinq scénarios pour comprendre les risques, du plus hypothétique au plus plausible.
- Professor of AI, Research Group Leader, UNSW
The most probable scenario is likely not what you think.
Anthropic’s latest model – Mythos – has spooked the world of finance. So it’s a good idea to keep your computer and banking apps updated.
If you’re one of CommBank’s 17 million customers, don’t panic. But given how well AI can now fake documents, all banks will need to rethink their security.
How might a super-intelligent machine take control? And what can be done to stop it happening?
Using AI to write search results is risky for Google, the internet, and the whole idea of ‘truth’
The Australian government has finally released a response to last year’s public consultation on the safe and responsible use of AI. Here’s what it entails.
The public release of the chatbot has led to a global conversation about the risks and benefits of AI – a conversation few people were having just a few years ago.
It has been an epic backstabbing scene worthy of the HBO drama Succession.
President Joe Biden has issued an executive order to regulate AI. The directives could help avoid AI doom – but they miss some key points.
The artificial intelligence boom means a multi-trillion dollar industry is coming into existence before our eyes. With great opportunity come great risks, as two important new Australian reports show.
AI is going to fundamentally transform how nations wage far. By failing to address it, the defence review leaves Australia unprepared for the future of war.
Users have reported some troubling experiences with the recently released search chatbot.
Will it give Microsoft a chance to overtake Google in the search space?
Ada Lovelace mengatakan bahwa komputer tidak dapat menciptakan. Namun, seabad kemudian, Alan Turing menunjukkan mesin memiliki kapasitas untuk menghasilkan sesuatu yang mengejutkan dan inovatif.
We’re facing a significant advance in AI using methods that are not described in scientific literature, and with datasets restricted to a single for-profit company.
Ada Lovelace said computers could not invent. But a century later, Alan Turing pointed out inventiveness in machines could be found in their capacity to produce surprising and innovative results.
The AI leans on a particular tactic for its success. If you knew what it was, beating it would get drastically easier.
The sentient, murderous humanoid robot is a complete fiction, and may never become reality. But that doesn’t mean we’re safe from autonomous weapons – they are already here.
Russia’s absurd claims about ‘smart’ landmines show it’s high time the world put limits on autonomous weapons.
Can software be an ‘inventor’? As courts wrestle with AI patent applications, the law must change to keep up.
We are connecting ourselves, our homes and our workplaces to lots of internet-enabled devices: smartwatches, smart lightbulbs, toasters, fridges … How to keep all their data private?
The collection of athlete data in professional sport has outpaced legal requirements and scientifically-proven benefit to athletes.
Basic research is best when it’s allowed to proceed on merit, rather than with political interference, says an open letter from 63 leading researchers protesting government interference in ARC grants.
As the perils and wonders of artificial intelligence begin to permeate our lives, the ‘IPCC report for AI’ calls for action from researchers and government to ensure a safe future.
Contact Toby for
- General
- Media request
- Speaking request
- Consulting / Advising
- Research collaboration
- Research supervision
- Location: Sydney, Australia
- Website
- X (formerly Twitter): @TobyWalsh
- Article Feed
- ORCID
- Joined
