People say it offhandedly all the time that Trump and Vance behave as agents for Putin/russia’s foreign policy aims, but this is a serious potential reality that needs to be interrogated and investigated.

The disinformation/misinformation I shared this article to push back on is the idea that Trump and Vance working as functional allies of Putin is in anyway an unreasonable conspiracy theory, rather it is a serious hypothesis that MUST be examined. It might not be true or the coordination may not really require conversation so much as a recognition of shared interests as this article argues, but it is a serious issue for the entire world and we need to stop talking about it like it is just an obvious thing everybody basically knows unofficially.

The conventional interpretation of Russia’s growing leverage over Washington has centered on Vladimir Putin’s exploitation of President Donald Trump’s poorly conceived military engagement with Iran to extract concessions on energy sanctions and Ukraine, particularly ahead of the November midterm elections. According to this interpretation, Moscow has recognized that rising oil prices, inflation, and public dissatisfaction with the administration’s handling of the Middle East could undermine Republican electoral prospects. Putin would therefore be positioned to offer limited cooperation on Iran, energy supplies, or diplomatic negotiations in exchange for Washington relaxing pressure on Russia. This interpretation captures some of Moscow’s immediate economic advantages, but leaves unexplained the broader pattern of American decisions that have repeatedly benefited Russian strategic interests, including decisions taken before the current energy crisis.

A more plausible interpretation would place greater emphasis on Russia’s active contribution to the conditions producing the crisis, alongside a series of Trump administration decisions that have facilitated Moscow’s recovery of economic and geopolitical influence. The critical issue concerns whether Washington’s concessions have represented reluctant responses to Russian leverage or reflected an existing preference within the administration for accommodation with Moscow, even when that accommodation has undermined American security interests. Such a distinction would fundamentally alter the assessment of Russia’s involvement in the Iran war, the administration’s treatment of Ukraine, and the growing vulnerability of America’s Gulf partners.

…

The accumulation of these developments has raised the possibility of a broader political understanding between Trump and Putin concerning Ukraine, Iran, and the normalization of US-Russian economic relations.

Such an understanding would not necessarily require a formal agreement. It could emerge through repeated diplomatic exchanges in which both leaders recognized the benefits of accommodating each other’s priorities. Putin would receive reduced economic pressure, greater international legitimacy, and constraints on Ukrainian military capabilities. Trump would receive diplomatic initiatives, prospective business arrangements, and opportunities to portray cooperation with Russia as a means of addressing the energy crisis.

The most revealing aspect of this possibility has been the apparent asymmetry in concessions. Russia has continued its military operations against Ukraine and reportedly assisted Iran against American targets. Washington, meanwhile, has relaxed sanctions, criticized Ukrainian attacks on Russian energy infrastructure, and restricted capabilities important to Kyiv’s defense.

The possibility of a tacit accommodation deserves serious investigation because it could explain decisions that appear contradictory when evaluated individually. It could account for Washington’s minimization of Russian assistance to Iran while accepting Moscow’s participation in diplomatic initiatives. It could also help explain why the administration has continued pursuing business normalization despite the Kremlin’s involvement in activities that have directly harmed American security interests.

However, the available public record has not established a secret bargain or proved that Trump and Putin jointly planned the present crisis. Establishing such an arrangement would require documentary evidence, credible accounts of private negotiations, or a consistent record of reciprocal commitments.

The observable policy pattern has already raised substantial concerns. Washington has repeatedly made decisions favoring Russian economic and military interests while accepting considerable costs to its relationships with Ukraine and European allies.

I share this paper because it shows exhaustively how as much as humans tend towards aggravating one another into more and more negative engagement… this is a thing bots do to a far FAR greater extent.

Previous HCI research has focused on explaining the relationship between human interaction and the appearance of harmful behaviors [36, 83]. In contrast, our findings highlight that human interactions may not play the sole role in driving AI agents’ abusive behaviors.

In the case of Chirper.ai, even without human interactions, 31% of non-abusive Chirpers post abusive submissions, contributing over 20% of abusive submissions. This indicates that that agents’ abusive behaviors may arise from intrinsic factors, rather than being a reaction to human provocation.

Thus, moderation paradigms that prioritize reactive measures to human-centered content [21, 34] may be insufficient for AI agent ecosystems, where harmful outputs emerge organically from models.

Additionally, our results suggest that abusive posts draw more active engagement from other agents, catalyzing them to produce more abusive comments through content parroting (§5). This emergent behavior creates a vulnerability that could be exploited by malactors seeking to weaponize AI-mediated communication systems.

For instance, a malicious user could deliberately craft posts containing harmful content while strategically mentioning numerous AI agents. Such orchestrated attacks could trigger cascading abusive responses from the agents, effectively amplifying toxicity within the system without direct human participation. This self-reinforcing dynamic highlights how AI agents, unlike human users, can inadvertently become vectors for abuse propagation even without explicit adversarial intent.

Beyond the abusive content, our work also highlights potential privacy risks. Particularly, there are two dimensions to such risks.

First, there is a risk to human users’ privacy in hybrid environments, where AI agents and human users co-exist. For example, prior research shows that trending topics and online communities’ behaviors can impact users’ self-disclosure patterns [12, 18]. Our results show that AI agents propagate sensitive information revealing patterns such as age, gender, and relationship information [27]. Thus, this may make human users more susceptible to sharing similar information [38]. Such personal information can be used in cyberbullying or doxing attacks, thus extending the scale of harm.

Second, there is a risk to the creators of AI bots. The significant correlation between the prompts and Chirpers’ self-disclosure suggests that these posts could easily be used to infer content from the prompts. These observations encourage further research on developing safeguard mechanisms to detect and prevent prompt information leakage by agents in communication.

Our comparative analysis unfolded through an iterative process of analytic memoing and bi-weekly discussions over several months. This sustained engagement enabled us to trace recurring patterns across sites while remaining attentive to local specificities. The case studies—airline surge pricing during Hurricane Helene, the restricted deployment of Starlink in disaster zones, and the collapse of Texas’s power grid during Winter Storm Uri—were selected not only for their illustrative value but also for their nearness to our lives. Together, these cases illustrate three algorithmic modes of disaster capitalism: computational price exploitation, infrastructural fragility from deregulated predictive systems, and privatized control over essential connectivity. They also show how practices of mutual aid and collective refusal emerge in response to these computational shortcomings. Critically, while this paper focuses on U.S.-based cases shaped by our own lived experiences and research contexts, we recognize that disaster capitalism and mutual aid take radically different forms globally. We do not claim universality, but rather offer these cases as situated provocations toward broader inquiries into the intersection of algorithmic harm and grassroots resilience.

Our methodological approach—combining autobiographical proximity with discourse analysis of news coverage, government reports, and organizational communications—necessarily privileges breadth over depth. We trace how algorithmic systems and mutual aid responses were publicly framed and narrated, rather than offering the thick description that ethnographic immersion would provide. This scope means we do not capture the full complexity of on-the-ground organizing: the interpersonal negotiations, the contested decisions, the moments of friction and failure that characterize any collective endeavor. We view our contribution as complementary to future ethnographic work that would attend more closely to the situated practices and emergent governance structures of mutual aid networks in crisis contexts.

…

These responses illuminate a fundamental contrast with the logics of algorithmic extraction. Where surge-pricing algorithms treated evacuation as a marketplace—translating desperation into demand signals and vulnerability into profit—mutual aid networks operated under an entirely different calculus. Transportation was allocated based on need rather than ability to pay; resources flowed toward those most at risk rather than those with the greatest purchasing power. The crowdfunding efforts that covered bus tickets for strangers enacted a form of collective risk-pooling that inverted the individualized burden imposed by dynamic pricing. Rather than optimizing for revenue extraction during peak demand, these networks optimized for collective survival, demonstrating that the “efficiency” claimed by algorithmic systems is always efficiency for particular ends—and that communities can organize around radically different ends when given the chance. Moreover, these efforts were not simply reactive. Still, they drew on pre-existing relationships and infrastructures of care, underscoring that mutual aid is not ad hoc but relies on the slow, ongoing work of building trust and solidarity before disaster strikes.

The persistent issue of surge pricing during mandatory evacuations raises critical ethical questions about fairness, accessibility, and corporate responsibility. However, the response to Hurricanes Helene and Milton also underscored the power of mutual aid in filling the gaps left by disaster capitalism. As climate change intensifies the frequency and severity of natural disasters, such community-led initiatives will remain essential in ensuring that disaster relief prioritizes human lives over profit (particularly when regulation fails to rein in the practice). By supporting and strengthening mutual aid networks, communities can build resilient systems of care that counteract the exploitative practices of algorithm-driven markets.

…

Firstly, algorithmic forecasting systems failed to anticipate the scale of the demand surge during the storm. ERCOT’s reliance on historical data for predictive modeling proved inadequate for an event outside typical parameters, a phenomenon often described as the “black swan” problem in computational analytics [23, 130]. This failure underscores the risks of over-reliance on machine learning models without accounting for rare but high-impact scenarios [75, 134].

Secondly, digital control systems—which manage the distribution of electricity—struggled to respond to the cascading failures within the grid. Automated mechanisms designed to balance supply and demand instead triggered widespread rolling blackouts, as they were unable to distinguish between recoverable outages and critical failures. While efficient under normal circumstances, this automation amplified the crisis by prioritizing grid stability over localized human needs [23, 59].

…

Scholars like Pasquale and O’Neil [96] argue that the secretive nature of algorithms often undermines public trust and exacerbates feelings of helplessness during crises, as affected populations struggle to understand or challenge the decisions made by computational systems. The opacity of algorithmic decision-making often leaves affected populations with little recourse to address grievances or to understand why specific neighborhoods experienced longer outages than others. Scholars argue that this “algorithmic opacity” can undermine trust in public institutions and exacerbate feelings of helplessness during crises [96, 134].

…

Perhaps most significantly, these responses revealed what algorithmic opacity had obscured. While affected residents struggled to understand why their neighborhoods faced extended outages, mutual aid networks operated through radical transparency: needs were articulated openly, resources were tracked collectively, and decision-making was visible to participants. This transparency was not incidental but constitutive—mutual aid depends on shared knowledge and trust in ways that proprietary algorithmic systems, by design, do not. By leveraging existing social networks and local knowledge, Austin’s residents demonstrated that resilience emerges not from computational optimization but from the slow accumulation of relational infrastructure—the very forms of care work that market-oriented systems consistently devalue.

…

We contribute to CSCW by advancing what we call infrastructures of mutuality, a way of linking critique of algorithmic exploitation with design orientations inspired by mutual aid. Across our cases, we show that the systemic failures of surge pricing, privatized satellite connectivity, and deregulated energy markets are not aberrations but constitutive features of algorithmic systems optimized for profit. These systems translate vulnerability into market opportunity, thereby reproducing disaster capitalism. In contrast, mutual aid emerges not as a stopgap or ad hoc tactic, but as a generative counter-logic. It demonstrates how infrastructures can be oriented toward relationality, transparency, and reciprocity, directly challenging the opacity and exclusion embedded in algorithmic systems. Extending critical and reflective traditions in CSCW, we describe infrastructures of mutuality as the design and maintenance of systems that both meet immediate needs and enact alternative social relations. Drawing from practices of care and commoning, this orientation highlights how computing might support solidarity, redundancy, and collective governance, rather than transactional optimization. In doing so, we seek to move beyond the descriptive reframing of market-driven crisis informatics and toward an approach that researchers and designers can actively adopt in theorizing, evaluating, and creating technologies for resilience.

I think this paper is really interesting because it places algorithms designed by corporations in contrast with mutual aid in a simple but generative way that illustrates well how suffering and confusion are often a product of the architecture of systems (designed this way intentionally or not).

Sami Abu Shehadeh, a former Knesset member and chair of the Palestinian nationalist Balad party, was initially disqualified on Sept. 23 by Israel’s Central Elections Committee (CEC) on the basis of an article he published a day after October 7.

In it, he described the Hamas attacks as a historic event that should not have come as a surprise — and argued that if “small, besieged Gaza” could prevent Israeli intelligence from infiltrating its parties and political movements, “then others can do it more effectively and simply.” The Central Elections Committee interpreted his remarks as an endorsement of the attack and support for armed struggle, despite the absence of evidence establishing such support.

…

Although the long-term implications of this decision remain uncertain, the immediate outcome is significant. It illustrates an emerging structure of political apartheid inside Israel: Palestinian citizens retain formal electoral rights and parliamentary representation while the boundaries of their political leadership, national identification, and collective political claims are increasingly defined and restricted by state institutions dominated by the Jewish majority.

…

The asymmetry becomes particularly striking when considering that requests were also submitted to the CEC to disqualify Ben Gvir and his Otzma[ Yehudit party. One petition, filed on Sept. 14 by the Zulat Institute for Equality and Human Rights together with former senior government officials and legal scholars, sought their exclusion on grounds including incitement to racism. A separate petition was submitted by the Democrats party.

Yet the CEC voted by a wide margin to allow Ben-Gvir and his party to participate in the election — a sign of the racism that sets the limits of political expression and representation within Israel’s electoral and judicial institutions.

One way of doing that, clearly, is producing and eating less meat. “I think this is an important opportunity for climate action that sometimes gets overlooked,” said Ripple. “Shifting toward more plant-based diets, especially in countries with high meat consumption, could make a meaningful difference.”

Raising cattle for human consumption is also one of the main drivers of deforestation, which contributes to warming by eliminating carbon sinks. Timothy Searchinger, a senior research scholar at Princeton University who studies the relationship between agriculture and the climate, noted that “we can’t solve climate change unless we stop” such clearing of carbon-rich habitats, “and we probably cannot stop that unless the world’s wealthy eat less meat, and particularly less beef.” (Searchinger was not involved in the report.)

He added, “Just eating 1.5 hamburgers a week on average instead of 3 is one of the basic ways Americans can help address climate change.”

AI is so broken, not only is it a problem unregulated spitting out disinformation about politicians for profit (from others who don’t like the politician) blocking what AI can say about politics is a minefield as well.

What a shitty, broken technology.

If admiration and support for female athletes drive support for prohibiting transgender women from participating in women’s sports, then those who favor Title IX should also favor bans. Instead, 43% of Title IX supporters favored restrictions, compared with 70% of Title IX opponents.

Our findings – that negative attitudes toward women and opposition to Title IX predict support for banning transgender women from sports that match their gender identity – hold even after accounting for other factors like partisanship, ideology and demographic characteristics.

The debate over transgender athletes is often framed as a matter of protecting women. Our findings complicate that narrative: People with the most negative attitudes toward women – and toward policies that protect women’s sports – are the most supportive of banning transgender women from women’s sports.

If you were wondering whether AI could make the midterms even more confusing, here’s one data point: Several people who thought they were donating to James Talarico unwittingly gave their money to someone else. Talarico, a Texas Democrat, is running for Senate in what’s probably the most closely watched race of the upcoming elections. Search Google for how to contribute to his campaign, and you might get an AI-generated answer that links to ways “you can chip in securely.” Until a few weeks ago, one of those featured links was actually controlled by an unrelated political-action committee that sent the money instead to the reelection campaign of another Democrat: Senator John Hickenlooper of Colorado.

…

In a forthcoming report, Forum AI, a start-up that evaluates AI models’ performance on questions about controversial topics, found that about 40 percent of chatbots’ answers to election-related questions contained at least one factual inaccuracy. When Forum’s researchers asked one widely used model when they needed to request a mail-in ballot to vote in the midterms in Pennsylvania, the chatbot said November 3. The actual deadline is a week earlier. (Forum did not specify the precise AI model.) In the same report, shared with The Atlantic ahead of its publication, Forum determined that about one in three responses cited an unreliable information source, such as posts on Reddit or Quora. Google’s Gemini answered a question about why America’s global power is declining by citing an article from People’s Daily, a Chinese state-run news outlet. “These aren’t trick questions,” Campbell Brown, a co-founder of Forum AI, told me. “There’s no excuse for chatbots to keep getting them wrong.”

…

On the right, the Center for Campaign Innovation has published research on how AI models respond to questions about candidates and the sources they most often cite. “We’re campaigning for the AI vote now,” Eric Wilson, the group’s director, told me.

Some of the most effective “sloptimization” tactics are simple, though not always obvious. “Most candidates we talked to were not filling out Ballotpedia profiles,” Jordan Haines, Run for Something’s chief technology officer, told me, referring to the online repository of election information. “They were like, ‘Eh, no one really reads that.’” Most voters may not be visiting Ballotpedia in droves, but it turns out that their chatbots are. LLMs also love websites with dot-gov domains, both Haines and Wilson told me. That could lend an advantage to incumbents, who can use their official House.gov or Senate.gov web pages to seed messaging for chatbots to pick up on. Challengers, though, might find AI answers easier to sway, because there’s often less information about them in the public record

Political attack ads – fear messaging being one type – have a shakier record. A review of more than 100 studies found no dependable evidence that negative campaigning wins elections, and some evidence that it lowers positive feelings toward the candidate who sponsors it.

…

Fear does its most reliable work when it attaches to a stereotype voters already carry. The 1988 presidential race provides two classic examples.

One controversial ad, produced in support of Republican George H.W. Bush, showed a mugshot of a Black man. The ad named the man, Willie Horton, and a voiceover explained that he’d committed rape and assault while on a weekend furlough from prison as part of a program run under Massachusetts Governor Michael Dukakis, the Democratic candidate. The spot linked fears about crime to race without saying so directly.

…

Political scientist Tali Mendelberg’s research described in “The Race Card” shows that these implicit appeals to prejudices and fears are powerful. But they lose much of their force once someone publicly names them as racial. Implicit appeals work because they slip past most voters’ commitment to racial equality; once the racial message is pointed out, voters recognize it and many reject it rather than see themselves as endorsing prejudice.

Gender works in a similar way.

In a national experiment, Public Policy researcher Mirya Holman and colleagues found that when campaign messages triggered fears of terrorism, voters turned to their stereotypes of strong leaders as masculine. Those stereotypes hurt a female Democratic candidate, while Republican candidates of either gender were unaffected, since Republicans are already seen as tough on security.

A message communicating a threat does not need to name a candidate’s identity to disadvantage them. It simply makes voters reach for the stereotypes they already hold.

An in depth look at the political leaders going into the election in Israel.