All desks

Aperture's new Upper West Side gallery opens with photographs by Diane Arbus, Dawoud Bey, Carrie Mae Weems and more. Free, with an amNY-listed run through November 1.

HiddenLayer’s checkout and toxicity-filter tests show how typed decisions still fail under manipulated input. The evidence establishes a failure mode, not a population attack rate.

A five-agent study separates cheap action selection from successful task delivery: recovery, auxiliary reasoning, and completion checks determine whether the savings survive.

John Rees examines how military authority and undisclosed commercial interests narrow Britain's debate about war and public spending.

An Upper East Side gallery opening tonight, free rock tomorrow and a free LIC DJ party Saturday, plus seven still-on reminders.

Thursday’s AI news is about making agents operationally dependable, with OpenCode 2.0 leading: its shared service and composable tools change how integrations work, but existing plugins need migration. Around it, Liquid’s downloadable agent …

A review of motivation research separates the felt cost of acting, the appeal of a reward and the difficulty of weighing both. Different obstacles call for different kinds of support.

1LittleCoder’s open-model routing experiment publishes real tradeoffs: lower estimated API cost, weaker full-set success, and a cheaper no-router baseline.

stern9 compares Jev and Claude on routing, phishing, and incident severity. The reusable artifact is the harness—not a general accuracy or calibration guarantee.

Vince Aletti on the magazine collaborations, commercial discipline, and selective postwar fantasy behind Avedon's photographic revolution.

An outage can empty a network's traffic chart without stopping its bots. Cloudflare follows the requests; HN questions the limits of that view.

Joseph Enochs on reversible agent actions, creative control in AI video, and finding enterprise problems worth solving—with important limits to the interview's optimistic claims.

A personalized news pipeline: deduplicate and filter in code, use Jev for inexpensive triage, and reserve an LLM for the ambiguous items and final digest.

Underdog is seeking federal protection from Tennessee's sports-wagering enforcement after an appeals ruling rejected similar arguments from Kalshi.

WIRED examines suspicious trading in long-open election markets and the limits of using volume as evidence of prediction-market credibility.

Louisiana granted a conditional license extension while seeking answers on AI, VIP treatment and advertising; the board has not found wrongdoing.

A Resolution Foundation report challenges diagnosis-based exclusions from Pip while proposing a different, contested route to savings.

An inspectable code-context selector publishes a paired relevance benchmark: much tighter context, but a lost required chunk and no demonstrated coding-success gain.

TypeSafe opens experimental image inputs, bounding boxes, and longer context. The important details are opt-in headers, version pinning, and an explicit absence check.

Reusing an open-source design is welcome. Crediting its authors should not disappear when an LLM helps build the successor.

Yiyun Li follows an overlooked musician from Pondicherry to Moscow, asking what fiction can rescue from history and grief.

How a state-built democracy of homeowners turned rising property values into a veto on the common good.

Hamel Husain's interview playbook: audit imperfect data, establish a baseline, run honest comparisons, challenge vague metrics, and explain product trade-offs.

Security specialists challenge the rogue-agent framing; the HN discussion separates model behavior, containment, and operator responsibility.

A real-session writeup shows where Jev was called, what its search results looked like, and which advertised judging features were only synthetic probes.

An inspectable Go library preserves Jev and Clef answers, applies explicit thresholds, and replays policy locally—with an important uncovered-question caveat.

Ben Taub revisits Kilis, where humanitarian storytelling, precarious freelance reporting, and the freedom to leave collided with civilians' inability to escape.

Claudia Verhoeven traces how prison correspondence became a neo-Nazi mythology—and why the differences between Manson and his admirers matter.

George W Brown's work made household relationships, chronic adversity and social support central to understanding relapse and depression.

Help restore natural areas, care for gardens, or maintain Randall's Island Park on Wednesday mornings through December 9. Registration is required.

Featherless’s inspectable scorer shares input processing across questions, constructs typed responses from token scores, and explicitly separates confidence from calibration.

TypeSafe documents unreliable counting, date comparisons, adversarial inputs, and option-order bias—and explains which work should stay in code.

Coding agents need more than source code: a way to replay a game, change its state, and check what actually happened.

A television crisis simulation turns arms-industry insiders into neutral experts and makes welfare cuts look like national survival.

French-Canadian loggers' involuntary startle responses became entertainment—and a revealing test of nineteenth-century medicine's class prejudices.

Free Midtown music this afternoon, Harlem jazz on Sunday, tonight’s Astoria gallery opening, and art, film and self-guided outings around NYC.

Wednesday centers on access and reliability: more developers can try an existing flagship model, while practical research exposes where agent infrastructure breaks. Gemini 4 Argon’s public rollout matters because teams can now evaluate it …

A case for programming without a business case: follow curiosity, build disposable experiments, and keep production obligations out of the playground.

Unmuzzled OX made a little magazine into a meeting place for poetry, conceptual art, queer life, and anti-war politics.

AP reports the first sentence in a college basketball fixing investigation, with prosecutors describing player recruitment, six-figure bribes and millions in fraudulent wagers.

An Alabama-Quassarte leadership dispute complicates Kalshi’s partnership claim as tribes challenge sports event contracts and their effect on gaming sovereignty.

Sportico reports a new Kalshi rule permitting sanctions for trading-related harassment, alongside a program to connect abusive social posts with betting accounts.

IndyDevDan’s Pi-agent lab separates bounded judgments from coding work—and its inspectable code exposes the limits of confidence gates and indirect file reads.

A tested Go proxy routes Claude Code by model and effort, while accounting for unfinished work, cache rebuilds, explicit overrides, and hosted-data boundaries.

A history of neighborhood socialist organizing: tenant struggles, political education, and Palestinian solidarity built trust that an election campaign alone could not.

When agents make code cheap, review and integration become the constraint. Anthropic's migration shows both the opportunity and the unanswered questions.

A feed reader's launch brought persistent bot load, not just readers. The discussion adds countermeasures and the cost of blocking legitimate visitors.

A live test finds real failures in 100 leasing-assistant conversations, but premature test design, brittle grading and a poor review interface undermine the workflow.

A practical framework for asking more of AI: bigger outcomes, deeper methods, and wider disciplines—with independent tests instead of confidence as the proof.

A Laravel/Filament plugin turns plain-English statements into row filters, with content-keyed caching, bounded requests, and an honest synthetic-label comparison.