Since January, Israel has repeatedly violated the terms of the ceasefire; it refused to withdraw its soldiers, it continued military operations (150 Palestinians were reportedly killed in Gaza during the “ceasefire”), and it blocked electricity and humanitarian aid from entering Gaza — a violation of international humanitarian law.

this week, he became the first Israeli leader to ever fire the head of Shin Bet, an intelligence agency in Israel that recently put blame on the Israeli government (and Netanyahu) for failing to act on warnings about the October 7 attack. With Netanyahu’s governing position secured and critical voices banished, the bombing started again in earnest.

while Hamas may be responsible for the horror it unleashed on Gaza with its October 7 attacks, I find it hard to dispute that — within the scope of the last two months — the primary fault for this deal collapsing lies with Israel.

The talking point from Netanyahu and Trump is simple: Hamas didn’t release the remaining hostages. But it’s also incomplete. Hamas did the most important thing it had committed to in phase one: It released all 33 hostages, as the deal called for, and came to the table to negotiate phase two.   Israel violated the agreement’s terms first by not meeting on the ceasefire’s 16th day to discuss the plans for phase two. Phase two was always going to be the sticking point, because it required an actual end to the war and Israel leaving Gaza.

Trump then began insisting on an extension of phase one, which was not in the text of the agreement. Then Israel broke the commitment to withdraw from the Philadelphi corridor. Then Israel broke its promise to continue aid while second stage talks were ongoing. Then Israel broke the promise to actually cease firing. Israel did all of this before Hamas balked on additional hostage releases.

Netanyahu has abandoned the hostages to extend his political life. From the early days of the war, this has been the story; it’s why he refused to end the war earlier, and it’s how he has survived this year and a half despite his political obituary being written on October 8. Once again, it is politically advantageous for him for the war to continue. Netanyahu needs approval for a budget before March 31, which he can’t get without support from the far-right wing of his party, which wants him to do exactly what he’s doing now.

Gazans have no ability to restrain or resist Hamas, a group that cares more about killing Israelis or pretending it may have a way to win this war than it does about protecting its own people. Israelis are at the whims of a leader who consistently ignores their pleas for a ceasefire, caring only about his own political survival.

we have the leaders of Hamas holding on to all they have, which are literal human bargaining chips, and the Israeli prime minister openly defying the desires of the hostages’ families. Meanwhile, the U.S. president openly muses about permanently vacating Palestinians from Gaza

we now know that as the price of computing equipment fell, new use cases emerged to fill the gap – which is why today my lightbulbs have semiconductors inside them, and I occasionally have to install firmware updates my doorbell.

surely the compute freed up by more efficient models will be used to train models even harder, and apply even more “brain power” to coming up with responses? Even if DeepSeek is dramatically more efficient, the logical thing to do will be to use the excess capacity to ensure the answers are even smarter.

ure, if DeepSeek heralds a new era of much leaner LLMs, it’s not great news in the short term if you’re a shareholder in Nvidia, Microsoft, Meta or Google.6 But if DeepSeek is the enormous breakthrough it appears, it just became even cheaper to train and use the most sophisticated models humans have so far built, by one or more orders of magnitude. Which is amazing news for big tech, because it means that AI usage is going to be even more ubiquitous.

The Intercept had a big story this week that is making the rounds, suggesting that “leaked” documents prove the DHS has been coordinating with tech companies to suppress information. The story has been immediately picked up by the usual suspects, claiming it reveals the “smoking gun” of how the Biden administration was abusing government power to censor them on social media.

As professor Kate Starbird notes, the Intercept article makes out like this was some nefarious secret meeting when it was actually a publicly announced meeting with public minutes, and part of the discussion was even on where the guardrails should be for the government so that it doesn’t go too far. Indeed, even though the public output of this meeting is available directly on the CISA website for anyone to download, The Intercept published a blurry draft version, making it seem more secret and nefarious. (Updated: to note that not all of the meeting minutes published by The Intercept were public: they include a couple of extra subcommittee minutes that are not on the CISA website, but which have nothing particularly of substance, and certainly nothing that supports the claims in the article. And all of the claims here stand: the committee is public, their meeting minutes are public, including summaries of the subcommittee efforts, even if not all the full subcommittee meeting minutes are public).

It includes four specific recommendations for how to deal with mis- and disinformation and none of them involve suppressing it. They all seem to be about responding to and countering such information by things like “broad public awareness campaigns,” “enhancing information literacy,” “providing informational resources,” “providing education frameworks,” “boosting authoritative sources,” and “rapid communication.” See a pattern? All of this is about providing information, which makes sense. Nothing about suppressing.

Alexios Mantzarlis, the founding director of the International Fact-Checking Network, worked closely with Meta as the company set up its partnerships. He took exception on Tuesday to Zuckerberg's statement that "the fact-checkers have just been too politically biased, and have destroyed more trust than they've created, especially in the US." What Zuckerberg called bias is a reflection of the fact that the right shares more misinformation from the left, said Mantzarlis, now the director of the Security, Trust, and Safety Initiative at Cornell Tech. "He chose to ignore research that shows that politically asymmetric interventions against misinformation can result from politically asymmetric sharing of misinformation," Mantzarlis said. "He chose to ignore that a large chunk of the content fact-checkers are flagging is likely not political in nature, but low-quality spammy clickbait that his platforms have commodified. He chose to ignore research that shows Community Notes users are very much motivated by partisan motives and tend to over-target their political opponents."

while Community Notes has shown some promise on X, a former Twitter executive reminded me today that volunteer content moderation has its limits. Community Notes rarely appear on content outside the United States, and often take longer to appear on viral posts than traditional fact checks. There is also little to no empirical evidence that Community Notes are effective at harm reduction. Another wrinkle: many Community Notes currently cite as evidence fact-checks created by the fact-checking organizations that Meta just canceled all funding for.

What Zuckerberg is saying is that it will now be up to users to do what automated systems were doing before — a giant step backward for a person who prides himself on having among the world's most advanced AI systems.

"I can't tell you how much harm comes from non-illegal but harmful content," a longtime former trust and safety employee at the company told me. The classifiers that the company is now switching off meaningfully reduced the spread of hate movements on Meta's platforms, they said. "This is not the climate change debate, or pro-life vs. pro-choice. This is degrading, horrible content that leads to violence and that has the intent to harm other people."

Crucially, this was never proven in court. And if Apple settle the case it never will be. Let’s think this through. For the accusation to be true, Apple would need to be recording those wake word audio snippets and transmitting them back to their servers for additional processing (likely true), but then they would need to be feeding those snippets in almost real time into a system which forwards them onto advertising partners who then feed that information into targeting networks such that next time you view an ad on your phone the information is available to help select the relevant ad.

Why would Apple do that? Especially given both their brand and reputation as a privacy-first company combined with the large amounts of product design and engineering work they’ve put into preventing apps from doing exactly this kind of thing by enforcing permission-based capabilities and ensuring a “microphone active” icon is available at all times when an app is listening in.

increasingly, a broad collection of conspiracy groups, extremist movements, political and commercial interests, and at times hostile states, coalesce around crises to further their agendas through online falsehoods, division and hate. They exploit social media moderation failures, gaming their algorithmic systems, and often produce dangerous real-world effects.

Some of the largest accounts sharing falsehoods about the hurricane response – including those with more than 2 million followers – have actively engaged with other forms of mis- and disinformation and hate. This includes anti-migrant conspiracies, false claims of electoral fraud, and antisemitic discourse around the so-called ‘Great Replacement.’ Their role as amplifiers here reveals how diverse groups converge on moments of crisis to co-opt the news cycle and launder their positions to a wider or mainstream audience.

Falsehoods around hurricane response have spawned credible threats and incitement to violence directed at the federal government – this includes calls to send militias to face down FEMA for the perceived denial of aid, and that individuals would “shoot” FEMA officials and the agency’s emergency responders.

The AI trust crisis 14th December 2023 Dropbox added some new AI features. In the past couple of days these have attracted a firestorm of criticism. Benj Edwards rounds it up in Dropbox spooks users with new AI features that send data to OpenAI when used. The key issue here is that people are worried that their private files on Dropbox are being passed to OpenAI to use as training data for their models—a claim that is strenuously denied by Dropbox. As far as I can tell, Dropbox built some sensible features—summarize on demand, “chat with your data” via Retrieval Augmented Generation—and did a moderately OK job of communicating how they work... but when it comes to data privacy and AI, a “moderately OK job” is a failing grade. Especially if you hold as much of people’s private data as Dropbox does! Two details in particular seem really important. Dropbox have an AI principles document which includes this: Customer trust and the privacy of their data are our foundation. We will not use customer data to train AI models without consent. They also have a checkbox in their settings that looks like this: Update: Some time between me publishing this article and four hours later, that link stopped working. I took that screenshot on my own account. It’s toggled “on”—but I never turned it on myself. Does that mean I’m marked as “consenting” to having my data used to train AI models? I don’t think so: I think this is a combination of confusing wording and the eternal vagueness of what the term “consent” means in a world where everyone agrees to the terms and conditions of everything without reading them. But a LOT of people have come to the conclusion that this means their private data—which they pay Dropbox to protect—is now being funneled into the OpenAI training abyss. People don’t believe OpenAI # Here’s copy from that Dropbox preference box, talking about their “third-party partners”—in this case OpenAI: Your data is never used to train their internal models, and is deleted from third-party servers within 30 days. It’s increasing clear to me like people simply don’t believe OpenAI when they’re told that data won’t be used for training. What’s really going on here is something deeper then: AI is facing a crisis of trust. I quipped on Twitter: “OpenAI are training on every piece of data they see, even when they say they aren’t” is the new “Facebook are showing you ads based on overhearing everything you say through your phone’s microphone” Here’s what I meant by that. Facebook don’t spy on you through your microphone # Have you heard the one about Facebook spying on you through your phone’s microphone and showing you ads based on what you’re talking about? This theory has been floating around for years. From a technical perspective it should be easy to disprove: Mobile phone operating systems don’t allow apps to invisibly access the microphone. Privacy researchers can audit communications between devices and Facebook to confirm if this is happening. Running high quality voice recognition like this at scale is extremely expensive—I had a conversation with a friend who works on server-based machine learning at Apple a few years ago who found the entire idea laughable. The non-technical reasons are even stronger: Facebook say they aren’t doing this. The risk to their reputation if they are caught in a lie is astronomical. As with many conspiracy theories, too many people would have to be “in the loop” and not blow the whistle. Facebook don’t need to do this: there are much, much cheaper and more effective ways to target ads at you than spying through your microphone. These methods have been working incredibly well for years. Facebook gets to show us thousands of ads a year. 99% of those don’t correlate in the slightest to anything we have said out loud. If you keep rolling the dice long enough, eventually a coincidence will strike. Here’s the thing though: none of these arguments matter. If you’ve ever experienced Facebook showing you an ad for something that you were talking about out-loud about moments earlier, you’ve already dismissed everything I just said. You have personally experienced anecdotal evidence which overrides all of my arguments here.

One consistent theme I’ve seen in conversations about this issue is that people are much more comfortable trusting their data to local models that run on their own devices than models hosted in the cloud. The good news is that local models are consistently both increasing in quality and shrinking in size.

Boreing said: "your story is inaccurate to the point of being false," though he did not specify a single inaccuracy, nor did he deny that the Daily Wire had sought and obtained a gag order on Owens at the same time they were publicly posturing as wanting a debate with her. The confirmation we obtained of all these facts is indisputable. Boreing added: "I’m sure you can appreciate how fraught a high profile break-up like this is. For that reason, we are trying to resolve our issues with Candace privately."It certainly seems true that the Daily Wire is attempting to achieve all of this "privately." Nonetheless, Ben Shapiro has constructed his very lucrative media brand and persona based on his supposed superiority in debating, a reputation cultivated largely as a result of numerous appearances at undergraduate schools around the country where he intrepidly engages with students who are often in their teens or early twenties. Both Shapiro and the Daily Wire have also predicated their collective media brand on an eagerness to engage in free and open debate with anyone, and to vehemently oppose any efforts to silence people, especially those in media, from expressing their political views.

To me it's pretty clearly the same targeted advertising available anywhere with the extra claim of using "voice data". It doesn't say what the voice data is or where it comes from. They could say that when people do google searches using Siri/OkGoogle/the microphone option on Google - it's information they would use in an anonymized way to target ads, or rather Google does on your behalf, and it's technically a derivative of voice data.

I'm skeptical this is what people might think it is. To be clear, I think most readers would interpret this as "your phone is surreptiously listening to you via your microphone." If that were true, then there would be telltale signs of resource draw. Handling rich audio data has practical costs, whether battery, CPU, network, memory, and/or disk; that data has to be stored, transmitted, and processed somehow. I've never seen analysis that shows that's happening. Not to mention this capability is beyond what audio capture APIs in Android and iOS offer, as far as I know.

Twitter is a cesspool of misinformation right now, with fake accounts, straight-up lies, confident-sounding assertions based on old videos, and people who spent 20 minutes Googling about something trying to correct reporters, academics and researchers who have spent years covering Israel, Palestine, and wars of all kinds.

An explosion at a hospital is not unlike the debate that broke out about precisely how many Israeli children were beheaded or burned alive after Hamas’s attack. Journalistically, it is important for us to get this story right. Morally, the finer details are in some ways not the most relevant. This explosion, whether it was the result of an Israeli rocket or not, has still left rescue crews collecting bodies and remains in a charred parking lot. And while everyone was arguing about who to blame this incident on, Israel did bomb a school run by the United Nations, killing six people. And the situation in Gaza is increasingly desperate.

They killed over 1,500 Hamas fighters in the initial battle. They have flattened dozens of buildings in Gaza — including schools, housing, and health centers. At least 19 journalists have been killed, a sign of how wide-reaching the violence has been. Israel has, in all likelihood, killed hundreds of Palestinian civilians — and the number may already be in the thousands. They have displaced one million people. Senior Hamas leaders are being killed left and right. They’re clashing with protesters in the West Bank, where the resulting violence has also killed people. They are rallying support from allies in Europe and the United States.That is the response. A lot of the commentary assumes Israel should pursue a strategy of uprooting and destroying Hamas, which implies that what we’ve seen so far is only the beginning. But as the U.S. learned in launching a full-scale attack to destroy al-Qaeda, what follows could well be worse. And as I wrote earlier this week, Israel should be careful not to repeat the same post-9/11 mistakes the U.S. made. Israel’s response has already been violent, and it’s been forceful, and it’s been severe.