Search the archive.

The Graphic Artists Guild position on AI generative technology is grounded on the fundamental rights of artists: credit for the work we do, our consent for the use of our work, and compensation for the tremendous value we bring. For our visual culture to continue to develop, the artists need to be respected and valued. These are individuals who have invested in their education and careers, who put up with harsh criticism and the risk of professional failure, and who have the tenacity and the drive to develop their craft and their careers. AI image generative technology would not be where it is today without having leveraged the labor of artists. They should be recognized, compensated, and protected.

The Guild’s position on generative AI is a live document that will be revised as the technology and policies developed around it evolve.

Autocrats start risky wars Three historically consistent patterns have emerged in the war in Ukraine. First, autocracies are more likely than democracies to start risky wars they go on to lose. Dictators are more willing to initiate high-risk wars because they know they can crack down on political opposition and remain in power if the fight goes badly. Iraq, for instance, launched two disastrous invasions of Iran and Kuwait. Even so, Saddam Hussein crushed internal uprisings to stay in power. Democratic leaders often, but not always, avoid these kinds of military failures, fearing the domestic electoral backlash of wars gone wrong. This is why democracies tend to win their wars — and why they wage shorter wars with fewer casualties. Like past authoritarians in Russia and elsewhere, Russian President Vladimir Putin has positioned himself domestically. He has steadily consolidated his hold on power over two decades, neutering the Russian parliament and political opponents, and destroying Russia’s independent media. Putin is now using that unchecked power to crush minor outbreaks of opposition, even from 7-year-old children. By eradicating the last vestiges of free speech and press in Russia, Putin may have felt comfortable taking on the risk of invading Ukraine. Dictators have to keep their militaries in check Second, like most dictators, Putin probably has some concerns about being overthrown by his own military. Dictators guard against this potential threat by centralizing military command authority and reducing the ability of lower-level commanders to take the initiative in battle. These moves may reduce an army’s ability to seize power in a crisis — but also undercut the military’s ability to defeat foreign foes. In its wars against Israel, the Egyptian leadership consistently restrained military leaders to lower the risk of an internal coup — but this crippled Egyptian fighting ability. Conversely, Israel’s willingness to grant command authority to lower-level officers has proven highly effective. Putin’s army today demonstrates the calcification and rigidity of a dictatorship. He appears unwilling to delegate decision-making autonomy to lower-level commanders, reducing military effectiveness. One result is that many high-level Russian officers lead from the front lines, where they have been killed in large numbers. Russia’s lower-ranked officers are not prepared or formally authorized to fill that leadership void. The Russian military’s inflexibility may explain other areas of poor performance. These include Russia’s curious vulnerability to Ukraine’s slow but lethal tank-killing drones, and the decision to drive a 40-mile-long Russian tank column into Ukraine along main roads exposed to ambush. Ukrainian troops, in contrast, are demonstrating bravery under fire combined with individual initiative — and inflicting surprising damage on Russian forces and supply lines. Dictators tend to ignore the truth Third, dictators often surround themselves with yes-men and political cronies, who deceive or remain silent rather than tell the unvarnished truth. Those in the inner circle typically reassure the leader that their vaunted troops will sweep aside the weak and corrupt enemy. In 1939, for instance, Joseph Stalin found out the hard way that invading Finland in winter was asking for trouble. In contrast, democratic leaders are more likely to have the benefit of robust debate inside and outside government. In 1990, President George H.W. Bush heard the hard truth from General Colin Powell that victory over Iraq would require a massive effort. Powell got the military commitment he said was necessary and delivered the most lopsided military victory in modern history. What about Putin? Every indication is that the Russian president is isolated and getting poor information. His arch-loyal defense minister, Sergei Shoigu, tasked with modernizing the Russian military over the last decade, promised Putin rapid victory — then apparently disappeared from view for nearly two weeks. Putin’s generals and intelligence chief reportedly refused to tell him the truth before the war: that years of Russian military reform had not made substantial progress, instead producing a “Potemkin military.” Actual combat is an unfailing truth-teller. Russian forces have suffered remarkable casualties and appear unable to execute basic battle requirements, such as implementing vehicle maintenance and providing troops with food and warm clothing.

Toward Synthesis What conclusions should be drawn from the spuriousness and causal arrow critiques, both with regards to academic scholarship and policy? Should political scientists recognize that the theory that democracy causes peace has been empirically disproven, meaning that the theory should be rejected? When considering whether or not to spread democracy, should policy-makers no longer consider the benefit that democratization might spread peace? There is enough evidence to draw the conclusion that joint democracy does cause peace, and that the dyadic democratic peace is a law. None of the spuriousness critiques, though intriguing, have sufficiently withstood scholarly rebuttals to justify dismissing as spurious the very strong correlation between joint democracy and peace, especially given the experimental, case study, and other quantitative observational work that provide support for different elements of the democratic peace argument. That said, the spuriousness critiques suggest possible modifications of the law of the democratic peace, such as perhaps that the democratic peace could be weaker in less-developed regions. Regarding the causal arrow thesis, though there is evidence that peace may cause democracy as well as democracy causing peace, the evidence is much weaker that peace causes democracy but not vice versa. Further, the claim that peace causes democracy but not the reverse contains theoretical inconsistencies. That said, it is of course conceivable that future studies may emerge that cast decisive doubt on the proposition that democracy causes peace. Data collection in international relations is never going to be as decisive in supporting or refuting theory as data collection in fields like physics or chemistry, where highly precise, often non-probabilistic theory permits point predictions that can be tested many times in controlled laboratory settings. It also will not be as decisive as data collection in the medical sciences, where theories are probabilistic but experiments can be conducted on thousands of subjects and repeated dozens of times. That said, the evidence that dyadic democracy causes peace is as strong as the evidence supporting essentially any theoretical proposition in international relations, other than relatively trivial propositions such as that adjacent states are more likely to fight each other than nonadjacent states. Echoing his 1989 assessment, Levy (, p. 587) remarked that in international relations “no one has identified a stronger empirical regularity” (Levy, , p. 587). That is, if the dyadic democratic peace is not a law, it’s as close to a law that we have in international relations, and probably as close to a law as we are ever going to see.

Recently, many scholars have highlighted the ways in which domestic institutions pacify interstate relations. This literature, commonly known as the Democratic Peace literature, argues that democracies tend not to fight (Maoz & Russett, 1993; Oneal et al., 1996; Bueno de Mesquita et al., 1999; Russett & Oneal, 2001). As one scholar puts it, ‘the absence of war between democracies comes as close as anything we have to an empirical law in international relations’ (Levy, 1989). This finding, having survived multiple operationalizations, empirical specifications, and econometric models, is now a standard control variable in the literature. How does democratic regime type operate to pacify international relations? One of the main mechanisms that scholars point to when explaining the democratic peace focuses on the role of audience costs in constraining leaders and helping them to signal their resolve (Fearon, 1994). Importantly, this audience cost mechanism relies on the assumption that there is uncertainty over each leader’s willingness to go to war. Proponents of the audience cost theory suggest that the greater ability that a leader’s relevant domestic audience has to punish leaders who lose disputes or escalate crises, the more one learns about the leader’s resolve (Fearon, 1995). Thus, the potential for domestic audiences to punish leaders for losing disputes or escalating crises reveals information that helps states to settle their disputes without going to war. Several scholars have tested and found evidence for this audience costs explanation (Gelpi, 2001; Tomz, 2007).2

Roger Lynch, CEO of magazine publisher Condé Nast, disagreed. "Fair use is to allow criticism, parody, scholarship, research, news reporting," he told the senators. "The law is clear when there is an adverse effect on the market for the copyrighted material ... Fair use is not intended to simply enrich tech companies that prefer not to pay."

There are other ways that tools like ChatGPT can eat into publishers' profits beyond reproducing their stories. Danielle Coffey, CEO of the News/Media Alliance trade association, noted that chatbots designed to crawl the web and act like a search engine, like Microsoft Bing or Perplexity, can summarize articles too.

I asked ChatGPT, a popular AI app, to answer this question: Can journalism survive AI?

The reason I gave it this task is because of a rather depressing report that just came out from the Brookings Institute on that very question.

Between many massive rounds of industry layoffs, a changing technological landscape, and increasingly eroding public trust in the news, the future of journalism is muddier than ever. And according to one former Big Tech executive? Big Tech's current project, generative AI, might put a nail in the media's coffin — or at the very least, fundamentally change what journalism is and how it's consumed.

In an opinion column for The Washington Post, ex-Googler Jim Albrecht — who until 2023 was the senior director for the search giant and major AI player's news ecosystem products division — argues that AI is the real "wolf" that threatens the business of journalism. Ever since the internet rose to prominence, the media industry has bickered with tech giants over things like search results and compensation models; but with AI now able to both paraphrase and deliver news, argues Albrecht, these qualms may already be mute.

These concerns are not as pressing if AGI is impossible, requires a paradigm shift away from deep learning, or some fundamental breakthrough in our understanding of intelligence. However, we may not be so lucky. There is a growing opinion that we may be able to achieve AGI with techniques similar to today’s deep learning systems with ever more compute. This scenario is called “prosaic AGI,” where we achieve AGI through prosaic (mundane) techniques. This possibility has begun to be taken seriously due to recent advances in large NLP language models pioneered by the transformers architecture. Computer scientists have observed that the training loss for models can decrease linearly with respect to additional compute [8]. As compute becomes cheaper as we follow Moore’s Law, we may observe a trend where existing models become more powerful without significant technological advances. Progress may accelerate by coupling prosaic advances with genuine computer science breakthroughs.

The key takeaway is that if this prosaic approach provides a plausible path to AGI, we may have less time to address many problems with deep learning that can pose risks to humans. Contemporary problems around discriminatory algorithms may become even more damaging as these models become more powerful and effective. A virtuous tool for good actors that can quickly become a weapon in the hands of bad actors. Furthermore, as we’ll explore in the next section, there are many open questions about how we’ll align AGI systems with human values.

One of the main missions of AI governance is value harmonisation for design, development, and deployment of AI technologies. As accountability regimes govern moral expectations and claims in social interactions, they promote alignment on specific values and beliefs (Dubnick, 2011). In AI, accountability fosters convergence towards values like fairness, inclusivity, and transparency. Under this governance mission, the goal of compliance should be emphasised. Value harmonisation engenders trust (Taddeo, 2017), which is another key mission of AI governance, one for which accountability plays a key role. Good governance ensures that expectations about the benefits of AI technologies are met. This requires understanding technological potential and risks, building an appropriate infrastructure to integrate AIs into specific environments, and developing safe, robust, and efficient AIs. Accountability rules, both proactively and reactively, introduce oversight and transparency mechanisms throughout the life cycle of AIs, and thus play a central role in fostering trust in AI. The greater the legal certainty on accountability schemes, the greater the public and private trust in AIs. AI market growth comes through legal certainty and widespread trust. Under this governance mission, oversight and report should be emphasised. To drive its impact on society, governance policies must distribute the costs and risks of AI. An efficient allocation of social costs should incentivise technological innovation, prevent or minimise damage, and ensure redress for victims. As accountability regimes assign duties and liabilities for social activities, they also distribute negative externalities. There will be different social effects depending on how the accountability burdens are distributed among stakeholders of AIs. The AI Act, for example, places most of the accountability burdens on developers and providers of high-risk AIs. 22 Under this governance mission, the focus should be on enforcement.

According to Yampolskiy (Yampolskiy 2016), the probability and seriousness of AI failures will increase with time. We estimate that they will reach their peak between the appearance of the first self-improving AI and the moment that an AI or group of AIs reach global power, and will later diminish, as late-stage AI halting seems to be a low-probability event.

AI is an extremely powerful and completely unpredictable technology, millions of times more powerful than nuclear weapons. Its existence could create multiple individual global risks, most of which we can't currently imagine. We present several dozen separate global risk scenarios connected with AI in this article, but it is likely that some of the most serious are not included. The sheer number of possible failure modes suggests that there are more to come.

Among the proliferating AI-related litigation, the New York Times filed a copyright infringement lawsuit against Microsoft and OpenAI. Along with other allegations, the New York Times claims that Microsoft and OpenAI are infringing copyright when they train their large language models (LLMs) on material copyrighted by the Times.

OpenAI has responded that “training AI models using publicly available internet materials is fair use, as supported by long-standing and widely accepted precedents.” In a blog post about the case, OpenAI cites the Library Copyright Alliance (LCA) position that “based on well-established precedent, the ingestion of copyrighted works to create large language models or other AI training databases generally is a fair use.” LCA explained this position in our submission to the US Copyright Office notice of inquiry on copyright and AI, and in the LCA Principles for Copyright and AI.

But our generation is gravely mistaken in its risk assessment. American and Russian nuclear arsenals remain alive and well today, and worse, they have been by nations such as India, Israel, North Korea and Pakistan. Many of us remain stubbornly attached to the theory of Mutually Assured Destruction, which argues the threat of nuclear retaliation against an attack provides an adequate safeguard against conflict escalation. Some scholars, including Kenneth Waltz, have gone so far as ing that we should allow nuclear weapons proliferation as a method of promoting peace. However, this deterrence-based approach does not take into account the possibility of s, or of , two very real menaces. Spreading nuclear weapons to less developed countries that may not have the means to safekeep them would drastically increase the chances of an unwanted detonation, potentially setting off a chain of catastrophes from which there will be no return. 

“In times of global uncertainty, the European Union has become a global point of reference for all those who believe in non-proliferation and multilateralism.

Our partners know where we stand. They know we are an honest broker, and we will always seek win-win solutions to preserve peace and security. And they know we have world-class expertise on non-proliferation – also thanks to the work of the European Network of Independent NonProliferation and Disarmament think-tanks. Independent analyses and recommendations are an essential contribution to our policy-making, so I am glad that the Council decided to extend our support to the network, also as a contribution to the implementation of the EU Strategy against proliferation of weapons of mass destruction.

A new transatlantic framework between the US and EU, comparable to Nato, is needed in order to address the challenges of the 21st century. Nato is currently performing indispensable service in supporting Ukraine against Russian aggression, as it did in the 1990s in ending the wars in Bosnia and Kosovo. But the alliance is consuming all the attention of leaders on both sides of the Atlantic at a time when we face issues that go beyond its mandate. Military security, economic security and prosperity are intertwined and must be co-ordinated, not kept in separate silos.

Consider the challenges we face: the rise of China; the existential threat of climate change; securing fragile supply chains and critical minerals; transformational technologies like artificial intelligence; looming nuclear threats from Iran and North Korea; and the task of rebuilding Ukraine. All of these demand greater US-EU co-operation than ever before.

THE WORLD'S MOST valuable and dominant internet companies are based in the US, but the nation’s unproductive lawmakers and business-friendly courts have effectively outsourced the regulation of tech giants to the EU. That has given tremendous power to Didier Reynders, the European commissioner for justice, who is in charge of crafting and enforcing laws that apply across the 27-nation bloc. After nearly four years on the job, he’s tired of hearing big talk from the US with little action.

Ahead of his latest round of biannual meetings with US officials, including attorney general Merrick Garland in Washington, DC, tomorrow, Reynders told WIRED why the US needs to finally step up, where a probe into ChatGPT is headed, and why he made contentious comments about one of the world’s most prominent privacy activists. His bicoastal tour began with a Waymo robotaxi ride through San Francisco (he gave it a rave review) and include meetings with Google and California’s privacy czar.

The Federation of European Publishers (FEP) was quick to praise the passage of new legislation by the European Union that, among its provisions, requires "general purpose AI companies" to respect copyright law and have policies in place to this effect.

FEP officials called the EU Artificial Intelligence (AI) Act, which passed on February 2, the "world’s first concrete regulation of AI," and said that the legislation seeks to "ensure the ethical and human-centric development of this technology and prevent abusive or illegal practices law, which also demands transparency about what data is being used in training the models."

On Friday, Italian regulators imposed a ban on generative AI tool ChatGPT with immediate effect while giving its creator, OpenAI, 20 days to address concerns about the way data is collected and processed under penalty of a fine of $21.7 million or up to 4% of annual revenues (whichever is greater).

There have been indications that other European regulators may swiftly follow suit. Reports suggest that France is conducting its own inquiry; Ireland has asked Italy for more details about the basis for the ban; and the German data commissioner has said that the same action could “in principle” be taken in Germany.

European parliamentarians are considering additions to the initial Artificial Intelligence (AI) Act’s draft lists of either prohibited or high-risk applications.

The AI Act’s main thrust is to require programmers working on high-risk applications to document, test and take other safety measures. Parliament is poised to rope in more applications than initially proposed, including broad categories such as systems “likely to influence democratic processes like elections,” or “General Purpose AIs” which can be integrated into different applications, such as OpenAI’s ChatGPT.

We come here to the fundamental moral dilemma. If, both biologically and psychologically, the need for survival is basic to man, and if survival is the precondition for any and all human achievements, and if no other rights make much sense without the premise of a right to life—then how will it be possible to honor and act upon the need for survival without, in the process, destroying everything in human beings which makes them worthy of survival. To put it more strongly, if the price of survival is human degradation, then there is no moral reason why an effort should be made to ensure that survival. It would be the Pyrrhic victory to end all Pyrrhic victories.

In 2017, scientists at Carnegie Mellon University shocked the gaming world when they programmed a computer to beat experts in a poker game called no-limit hold ’em. People assumed a poker player’s intuition and creative thinking would give him or her the competitive edge. Yet by playing 24 trillion hands of poker every second for two months, the computer “taught” itself an unbeatable strategy.

Many people fear such events. It’s not just the potential job losses. If artificial intelligence (AI) can do everything better than a human being can, then human endeavor is pointless and human beings are valueless.