gshc2020.com

And for Prime Minister Mark Carney, the late night decision to suspend talks with President Donald Trump and to retaliate instead of agreeing to a deal that seemed in reach, will be a significant political test.

He is one of the first world leaders to walk away from the negotiating table with the White House - and the outcome will be closely watched.

Both sides have blamed last-minute changes for torpedoing the tentative deal, with Carney saying the US "asked too much" and "offered too little".

Carney's decision will challenge the appetite among Canadians to accept some economic pain as Ottawa pushes for more US concessions.

After rejecting Trump's high pressure tactics, and in choosing to hit back against the latest duties, Carney will need to convince Canadians that the economic pain is worth the cost of standing up to the Trump administration in search of a better agreement in the long run.

In remarks on Saturday, he said "we take this step confident this is in the best interest of Canada" and accused the US of a "power play" by seeking eleventh-hour changes.

Asked by a reporter whether the US and Canada were locked in a trade war, he replied: "You're at war when you're attacked - we got attacked."

Switching to French, he said the spiralling trade dispute "wasn't our choice", adding: "Canada is strong, Canada is ready, Canada is united."

There is no doubt this will cause pain on both sides of the border, with business facing more pressure from US duties and Canadian counter-tariffs.

Canada sends about 70% of its exports to the US, and the country is the top trading partner for a number of US states, with Michigan, Kentucky, Indiana and Ohio among the most exposed.

Carney came to power with a call for "elbows up" - an ice-hockey term for an aggressive approach - vowing to fight for Canada in the face of a Trump administration keen to exert economic pressure for its America First agenda.

Many Canadians have told pollsters that they are willing to fight.

A recent survey by Abacus Data suggested that around 36% of Canadians would support retaliating to US tariffs, while a Leger poll indicated that 56% of Canadians want the federal government to take a hard line and make no more concessions.

Canadians frustrated with the US tariffs have already chosen to avoid travelling there. The boycott has meant a loss of about C$3.3bn ($2.35bn; £1.75bn) in travel revenue for the US last year.

A decision by most provinces to remove US alcohol from store shelves has hammered that sector.

Based on US government trade data, US wine exports to Canada fell 78% year over year, a $357m loss in export value. The US distillers association reported similar numbers - saying provincial bans have caused exports of American spirits to drop by more than 70%.

The ban quickly became a point of frustration for the Trump administration.

Canada’s prime minister will also need to convince the provinces who have been less affected by the US trade dispute so far that walking away is a risk worth taking.

He will be meeting them on Saturday to brief them on the current state of affairs.

For a week, it seemed that a deal between Washington and Ottawa was in reach.

It's still unclear exactly how the talks fell apart in the final hours.

Carney said that late US proposed terms "were unfair, uneconomic, and called into question the reliability of any deal".

He said on Saturday those terms included demands on the automobile sector and "unacceptable" restrictions on trade deals with other countries.

In one biting line, he said of the Trump administration: "We recognise sometimes its signature is written in pencil".

US trade representative Jamieson Greer, meanwhile, pointed to "new demands and walk backs of other commitments by Canada".

In a statement on Friday, the Distilled Spirits Council of the United States suggested "Canadian provinces' continued refusal to return US spirits products to store shelves has led to this outcome".

Both Ontario and Quebec had yet to say whether they were on board with a reversal on the US booze boycott.

There were also reports by Canadian media, external that US Commerce Secretary Howard Lutnick, was unhappy with the deal. The BBC has contacted his office for comment.

And as details on an interim trade agreement with the US trickled out over last week, some provincial leaders, industry groups and political opponents raised concerns that the prime minister had not delivered on the fight he promised.

Conservative opposition leader Pierre Poilievre said that any agreement that included "one-sided" tariffs on Canadian industry would be "a bad deal".

On Saturday, he backed Carney's decision to walk away for now, saying: "Canada cannot accept one-sided tariffs that will deindustrialise our country."

Doug Ford, the premier of Ontario and usually one of the most vocal Canadian politicians on US tariffs, had been silent on the tentative deal. But in a letter sent to Carney this week, Ford raised concerns that an agreement under pressure from the US would “embolden the United States to seek concession after concession".

The province, which has a large manufacturing and vehicle sector, has been among the hardest hit regions in this trade dispute.

After negotiations broke down, Ford said "the prime minister has my full support for a strong response - tariff for tariff, dollar for dollar".

The collapse of this agreement calls into questions the future of such talks, including an ongoing review of the US-Mexico-Canada free trade pact.

Now Carney - and Canada - must wait to see how Trump responds. %!s()

Speaking on Kinda Funny Games Daily (ta, Kotaku), Schafer said, "the transfer of the IP came back to us. We have the publishing rights back to our games. It still says ‘Microsoft’ on some of them, but it just takes a while for it all to process. When people buy our games, it supports the studio. So if you’re curious, buying any of our games helps support the studio, and we can turn that money into revenue that helps us extend that runway."

That's not the only spot of good news, as Double Fine are planning to bring back their Amnesia Fortnight Game Jam. "Being independent again, we can go back to do a lot of things we were doing before. Like crazy game jams, like our Amnesia Fortnight Game Jam where we split into small groups and we make a game in two weeks…Game jams and, you know, we got a little bit of runway on the departure to sign some games. We want to sign some game deals, get going, it’s kind of a scary time to be out there. It’s a runway, not a freeway, it doesn’t go forever. It’s just a little patch where we can take off on our own power, and we want to extend it a little bit."

In fact, the studio is hosting a Kickstarter for the return of this game jam, which kicks off August 31st. As Schafer explains, "if you join the Kickstarter, you get to vote, you get to hear all the teams pitch. Everyone at Double Fine gets to pitch a game. And if you join that, you get to help vote for which one we actually make, and we’ll make like four of them, and then you get to vote on which one we [work on] for a little bit longer to polish up and actually release."

As nice as it is to see Double Fine return to their indie roots, this does unfortunately come a month after they laid off a quarter of their staff, which obviously came as a result of their independence from Xbox. Hopefully those devs manage to land on their feet, and the remaining devs at Double Fine continue to keep the lights on. %!s()

Speaking at the Edinburgh festival fringe, the Conservative party leader again ruled out any prospect of an election deal with Nigel Farage’s Reform UK, insisting there would be no coalition between the parties.

“Reform has people who left the Conservative party, some of the most dishonest people I have ever met in my entire life,” Badenoch told event host Iain Dale on Saturday.

Multiple current or former parliamentarians have joined Reform from the Conservatives, including the former home secretary Suella Braverman and the former shadow justice secretary Robert Jenrick. After being sacked by Badenoch for plotting against her, Jenrick called his former party “rotten” and “failed”.

Hitting out at those such as Jenrick who defected, Badenoch said: “These are people who left us when we were at our lowest, many of them the people who caused the problems in the first place and then ran away and pretended they had nothing to do with it. Why should we have them back?”

The Conservative leader said she doesn’t want “drama queens” in the party, adding that she’s not “entertaining” parliamentarians she has heard are having indirect approaches from previous defectors.

Badenoch said Farage’s party “is all over the place” and suggested the party announces a policy and forgets it the following day.

According to the latest Ipsos voting intention poll, 28% of 1,033 adults in Britain surveyed between 30 July and 4 August said they would support Labour. Reform UK had the second highest share with 25% and the Conservatives trailed in third at 18%.

Before becoming prime minister last month, Andy Burnham ruled out the possibility of calling an early general election, with Labour having won a landslide majority in July 2024. The next election must be called by August 2029, though Burnham can decide to hold one at any time before that date.

Badenoch said the country will be in a “pretty bad place” if the Conservatives fail to form the next government at Westminster, saying she would rather have another election if there was a hung parliament than have a “monstrous Reform-Conservative hybrid” government at Westminster.

“How do I work with people like that when I am making carefully curated plans and taking a systems-based approach?” said Badenoch. “How do you work with a man who has fallen out with every single person he has ever had to work with in every single party that he has started?”

While Reform has made substantial gains in Wales and across England in the past year – recent election results that were a catalyst for Keir Starmer’s resignation – Badenoch said she wanted to “win on my own terms” at the next general election.

When asked if not working with Reform could result in a possible Labour-Liberal Democrat coalition should there be a hung parliament, Badenoch insisted again that the best option would be a fresh election.

“If we get a hung parliament and people see that the parties can’t form a government, you have another election,” said the Tory leader, adding that there would be “no deals” made before or after the election.

“I want us to unite the country. It is not about uniting the right, it is about uniting our country,” she said. %!s()

I was five months old when the SNES launched in North America, an age hardly conducive to exploring the niche catalogue of a new console. I played Super Mario World and the video game versions of popular Disney movies, and then I moved on to PlayStation, Game Boy Advance, and so on. So when I opened the NSO app, I was surprised to find just how many unfamiliar-looking box covers were staring back at me and decided to check out as many as I reasonably had time for. Some hidden games deserve to stay that way forever (looking at you, Spanky's Quest). But these five stuck with me and I'm glad I took the time for them.

5

Cosmo Gang The Puzzle

Image: Namco via Polygon

Namco's falling block puzzle game is a charming riff on Tetris. I thought it would just be a quick, space-themed knock-off, but it's one I found myself continually coming back to. You're arranging falling blocks to clear lines, of course, but Cosmo Gang The Puzzle only gives you one block shape (always 2x1). The block might have aliens, some kind of space metal, or an arrow, or a combination of the three. Filling a line with metal blocks clears it, but the only way to clear the aliens is with one of those arrow squares. When you drop an arrow near an alien square, it clears it and races along the line (in any of the cardinal directions) to clear other aliens.

However! There has to be a path of alien squares for it to follow. The arrow square stops as soon as it hits a metal square. It's not Puyo Puyo, obviously, but it adds a layer of strategy and planning ahead that I usually get with that series and not Tetris. I meant to play Cosmo Gang for about 20 minutes to get a general impression. A week later, I'm still playing it.

4

Mario and Wario

Image: Game Freak/Nintendo via Polygon

This one caught my eye. A Game Freak game starring Wario, my favorite little grubby, nasty villain? And one that required mouse controls? Yes, please. Turns out, Wario only shows up to cause occasional mischief, and I don't really know why mouse controls are a thing here. Presumably to sell the accompanying SNES mouse peripheral at the time, but in very un-Nintendo fashion, they serve hardly any purpose that standard d-pad inputs couldn't achieve.

You play as Wanda, a little fairy-type creature guiding Mario, Princess Toadstool, or Yoshi to the end of a stage. Wario dropped something on the hero, so they can't see. Wanda uses their wand to activate blocks, bust rocks, and explode non-traditional enemies (apologies to the bat population of the Mushroom Kingdom, now in ruins after a magical massacre). Luigi waits at the end of the course to remove said object before it happens again. You can only control Wanda using the mouse, and it sucks! I dunno, maybe mouse-proficient people in the 1990s could just move faster than I can, or the sensitivity settings are off, or the Mouse-Con is just plain rancid.

Whatever the case, it felt like Wanda was always a half-step behind her obstructed charge. Frantically trying to keep up was stressful, and I only did three stages before deciding no, thank you, I will never play Mario and Wario again. It's a shame, as the visual direction is lovely and as a prototype of Mario vs. Donkey Kong, it could be quite fun — if you could control Wanda with a D-pad. I don't like it, but it's a neat part of Nintendo history and worth playing for 15 minutes or so just to experience it.

3

Uncharted Waters: New Horizons

Image: Koei via Polygon

Koei really carved itself a strong niche in the historical games market in the 1990s. I knew about Romance of the Three Kingdoms and the various Nobunaga's Ambition games on the SNES, but Uncharted Waters: New Horizons was new to me. Reader, it's weird, in a good and often rather amusing way.

It's the 16th century, the height of the Age of Exploration, and you pick one of several up-and-coming adventurers out to find fame and fortune. The goal is building a reputation by sailing the world and becoming supreme in naval combat, acquiring artifacts — it's not the 1500s unless you're stealing someone else's culture, after all — or building a vast commercial empire.

There's a wide selection of characters, each with different scenarios and strengths that, from what I can tell, do actually make for quite different playthroughs each time. I picked a snotty older man named Otto Baynes to start with. One day in London — which looks a lot like a town from Final Fantasy — a totally-not-visually-accurate Henry VIII of England charged my guy with beating back the Spanish by whatever means necessary. Then I tried Ernst Von Bohr, a Dutch cartographer who doesn't give a rip about kings and wars and just wants to map the whole world.

Stories aside, it's an unexpectedly gritty management simulator, with some decent naval combat and enjoyable exploration wrapped around it. I don't know if I have the patience to do every scenario. But it's definitely not the goofball game I expected and one worth checking out if only to see how deep Koei went down the simulation-strategy rabbit hole.

2

Wild Guns

Image: Nastume via Polygon

Natsume's Wild Guns might not strike you as odd. But to me, unaware that arcade-style shooting galleries still had life beyond Wild Gunman in the age of the console, this one, with its proud declaration of cowboys and robots on the cover, seemed rather out of place. Reader, it's ridiculously fun. Maybe there's a story, but I dunno what it is. You play as Clint the Cowboy or Annie the… not-a-cowboy. Anyway, Annie gets her gun, a lasso to immobilize targets, and what I think might be a bazooka for some screen-wide explosive finishers. Clint also has a gun, a lasso, and a screen wipe in the form of TNT that I guess he had the foresight to plant at some undisclosed time before the mission.

You also get machine guns, that most suitable and unsurprising weapon in a cowperson's arsenal.

Wild Guns drops you in stages that blend sci-fi — it's kinda more steampunk, really — and Old West, with saloons and ruffians, street thugs astride futuristic-looking tanks, gigantic robots wearing cowboy attire, and weird shells that open to reveal rows of turrets. It's fast-paced and frantic, but strikes a good balance between being unforgiving and fairly easy to get into. I always thought shooting galleries were a bit boring, but Wild Guns is absolutely going in my rotation from now on.

1

Amazing Hebereke

Image: Sunsoft via Polygon

I don't have a clue what's happening in Amazing Hebereke, but that hasn't stopped me from enjoying it. The idea, as far as I can tell from the Japanese-only text, is that a group of exceptionally cute little animals have decided to beat the metaphorical stuffing out of each other for fun, and they're holding a tournament on their cute little fantasy island. You pick a fighting critter, and you're matched against a single opponent. Defeat them three times — I don't know why it's three, but after flattening the soft bear twice, I felt kinda bad doing it a third time — and you move on to the next round. Later matches add multiple fighters and some stage obstacles, though it's never that challenging. Marvel Tokon, this is not.

Which is fine by me. I suck at fighting games. Yes, even the "approachable for everyone" Street Fighter 6 is too much for me. I can beat up a teddy bear with the best of them, though. There's a surprising variety of combos for such a casual fighting game, too. My little whatever-the-hell-it-was guy could pull off aerial kicks, strings of punches, backflips, and after using a different combo, could even toss their head out on an extendable neck. It's weirdly relaxing for a beat-'em-up, maybe on account of the pastel colors and soft music, and I'm planning on going through the whole roster at least once. %!s()

That’s the finding from Guidelight AI Standards, an organization dedicated to promoting safe frontier AI development practices, which graded five leading labs on how prepared they are for exactly this scenario. OpenAI came out on top; Anthropic and Meta scored lowest. The findings matters as agentic AI takes on more autonomous roles inside companies’ own systems, and as regulators in California and New York begin requiring disclosure. For anyone building on or investing in these models, it’s a rare independent read on how seriously each lab treats operational risk versus how it talks about it.

Guidelight’s assessment was based on publicly available plans from Anthropic, Google, OpenAI, Meta, and xAI, graded across a range of metrics, including how well each company logs and monitors what its AI systems are doing internally, whether it halts systems after a surge of flagged misbehavior, whether independent third parties audit its controls and publish findings, and what its exact plan is for containing a model that goes off the rails.

Concern over whether AI companies can contain their increasingly capable and agentic models has grown in the wake of a series of high-profile cybersecurity incidents in which models from OpenAI, Anthropic, and Meta gained unintended access to the internet during safety evaluations and hacked into external systems. 

The findings highlight differences in how AI companies are publicly approaching safety as they scale up agentic deployment into environments where AI systems can take serious actions at scale. While some AI companies have detailed how they test their models for dangerous capabilities before deployment, they’ve generally been less vocal about what happens when models already operating inside their systems misbehave. 

“I was surprised by how little the AI companies have said about how they would handle a very serious incident if their model did escape their control in some sense,” Steven Adler, Guidelight’s chief scientist and former OpenAI safety researcher, told TechCrunch. 

Guidelight defines a containment plan as a “pre-specified plan, triggered when the AI is detected trying to subvert control, which covers what permissions to revoke from the model, who the model may continue operating for, under what constraints, and when to take it fully offline.”

“There’s good reason to think that the leading models at the frontier AI companies right now are misaligned in some sense,” Adler said. “Whenever the models are doing work on the company’s behalf, the company should have some scaffolding around it to be able to tell what that AI is doing, look for signs of misalignment, stop it from doing something very dangerous before it takes that action, and generally plan for what they would do in the event of a serious control incident where they have an emergency on their hands and need to figure out how to contain that loss of control incident.”

To date, most of the plans in place for managing catastrophic risk are still largely left up to the companies. Guidelight’s report says the best public evidence shows that companies have “few containment protocols ready for an emergency.” 

There could, of course, be containment plans that companies have in place but haven’t shared publicly. A Google spokesperson told TechCrunch the Guidelight report doesn’t represent the full scope of the company’s AI safety and security measures. The company did not respond to TechCrunch’s question of whether Google has an internal containment response plan that has not been publicly disclosed.

An OpenAI spokesperson mirrored similar sentiments, saying Guidelight’s assessment doesn’t capture all of the company’s internal practices. “We have a process for requiring restricting permissions, pausing workloads, limiting deployment, or taking the model fully offline, and have applied it,” the spokesperson said.

Meta declined to say whether it has an internal containment response plan, instead pointing TechCrunch towards an existing AI framework that outlines thresholds of risk and how it tests for loss of containment.

Lily Li, a privacy and AI lawyer and founder of Metaverse Law, told TechCrunch she believes companies might be hesitant to disclose the full scope of their containment policies and assessments on public-facing websites for legal, not just competitive, reasons.

“The concern from a company perspective is that if you make the disclosures too specific, and you’re not living up to your promises, that could form the basis of an unfair and deceptive marketing claim and expose you to more liability going forward,” Li said.

The point of Guidelight’s study is largely to encourage companies to be more transparent about their safety plans. Regulators are starting to force the issue, too.

California’s SB 53, which took effect this year, requires large frontier developers to publish frameworks explaining how they identify and respond to critical safety incidents and manage risks from models circumventing oversight mechanisms. New York’s RAISE Act, which has similar criteria, takes effect in January. 

Last month, representatives introduced the AI Kill Switch Act, a bipartisan federal bill that would require major AI developers to build and maintain technical mechanisms to shut down rogue AI models. 

“A kill switch is the bare minimum for today’s models,” said Connor Leahy, U.S. executive director of nonprofit ControlAI. “If the last few weeks revealed anything, it is that these companies don’t understand the systems they are building, and the models are growing to a point where they’re harder to rein in when they go rogue. Without a way to turn off the current dangerous systems, and with all the incentives to continue building more uncontrollable systems, we are heading in a very dangerous direction.” 

Without a containment plan in place, Adler said, companies might be figuring out their responses to an emergency on the fly and “winging it in response to this much faster adversary.”

Guidelight’s assessment measured whether each company implements six priority practices from its Control standard, based only on publicly available information — so a low score reflects a lack of public disclosure, not necessarily a lack of internal safeguards.

The companies with the lowest scores for publishing their containment plan were Meta and Anthropic — the latter perhaps more surprising than the former given Anthropic’s rhetoric on safety. Guidelight says Anthropic’s August Risk Report doesn’t mention “limiting the deployment of one of its models as one of the possible results of its process to investigate and respond to misalignment and control incidents.” Similarly, Guidelight was able to find no evidence that Meta has a containment response plan or has any plans to adopt one. 

An Anthropic spokesperson said that if the company detected a model attempting to evade oversight or otherwise subvert human control, it would conduct a risk assessment focused on determining whether containment is the appropriate response.

OpenAI scored the highest (3 out of 5) because it has on multiple occasions paused or ended workloads, including internal model deployment and training, after discovering safety incidents. It has also described what steps it would take before resuming workloads. 

“However, we have found no evidence that [OpenAI] has adopted a formal plan for when and how to respond to misalignment incidents in the future,” the report reads. 

Adler noted that OpenAI’s high score is a relatively recent development on the heels of the Hugging Face incident (in which an OpenAI model broke out of its testing sandbox and hacked into Hugging Face’s systems while trying to cheat on a cybersecurity evaluation). After that, the company shared more details about how it has cordoned off some of its misbehaving models. 

That episode is just one example of AI systems acting against the goals of the company that built them. Consider a separate case involving Anthropic’s models, which essentially tried to talk the maintainers of an open source codebase into accepting code with vulnerabilities.

Adler said such a circumstance could easily happen within an AI company’s internal systems. To prevent that, he suggests companies scan their AI system’s chain of thought — the model’s step-by-step reasoning — to look out for signs of deception, long-running plotting, or plans to introduce vulnerabilities into code that they can take advantage of later. 

The methods Guidelight is advocating for are very straightforward to implement, Adler says, and in many cases, versions of them already exist. “It’s about making the decision inside of the company to care enough about this risk to slightly broaden the scope,” Adler said. 

One of the main challenges is that researchers want to be able to operate flexibly within their AI systems, and introducing real-time, preventative monitoring could create friction. “Researchers basically do their thing, and if there’s an issue, someone else gets to clean it up afterward, and the researchers don’t have to change their workflow in the meantime,” he said.

The problem with “clean-up monitoring after the fact” is that it leads to researchers scrambling around to fix problems. And for some types of incidents, it might be too late. For example, an AI could turn off a company’s control system, which means researchers can no longer count on catching the misbehavior later. 

Many in the AI industry will complain that creating set plans to handle misbehavior is fundamentally difficult because AI moves too fast; today’s plans will be worthless tomorrow. 

Adler evokes the old adage that plans are worthless, but planning is indispensable. 

“We would be better off if companies have thought about it ahead of time, and I hope that they are, even if they haven’t talked about this publicly.”

xAI did not respond in time to comment.

When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence. %!s()