> OpenAI nearly doubled its federal lobbying expenditure to a record $2.22mn in the first half of 2026, compared to last year, while Anthropic nearly tripled its spending to $3.53mn, according to federal disclosures.
Never ceases to amaze me how cheap lobbying is. That's pocket change for these companies.
Bun’s Rust rewrite shipped in Claude Code over a month ago and barely anyone noticed. Claude Code is widely used. The Rust rewrite is going well overall.
In the Bun v1.4 video, I promised a certain number of newly passing Node.js tests were added to force us to improve compatibility, and that number is not true yet. The release is delayed until it is true. The PRs to make it true are up but not merged yet. Most likely next Tuesday we’ll do the release of 1.4.
> why would you clean them when you're about to sell the house?
So they look good and the potential buyer sees something that looks to be in good condition and working well.
I've limited sympathy for the publishers.
It pisses me off to reflect that they can sit on works until copyright expires, keeping them out of print. There's no real need for any of these so-called rare books to be rare while they're under copyright.
And related to this, the books that are in print are mostly only in print in the shittiest way. I often see well-made books from the 17th or 18th centuries which are still in good nick. It's ridiculous that in the 21st century, publication standards have fallen to the point where for most works a disposable format is the only type available - where no amount of money could buy a truly decent hardback copy.
If we have to have copyright laws, I'd like to see two changes to them.
When a publisher has no incentive to keep an edition in print, it should be available to any other publisher to print, without compensation to the original publisher, and with renegotiated royalties for the author.
And if the publisher keeps a book in print - but only in bestseller-grade materials, bogroll paper that furrows in any humidity and perfect binding that molts its pages a couple of dry seasons later - and if it refuses to print a durable hardback copy with signatures, good paper and decent print - something that will still be readable in several generations' time - any other publisher keen to have a crack at it should be able to, again without any compensation for the original publisher, though perhaps in this case, with matching royalties for the author.
The EU told them they can not by default prefer their own maps product when linking from their search product, they have to allow the user to choose.
They could have simply added a selector when a user first clicks on the maps preview in the search result, and then remembered it on device or across that user's account.
But of course, then the user could choose a competitor's product and Google would have to honor it. That would be horrible, so instead they just made the user experience worse for everyone and elegantly made people blame the EU.
I co-wrote a border search guide for EFF some years ago. I was very interested in finding clever technical approaches but I later ended up feeling that I hadn't given enough thought to the overall threat model questions (even though the guide did address them, perhaps even somewhat usefully).
The big picture problem is that the agents performing the searches have an enormous amount of power in terms of potentially seizing devices and potentially denying entry for non-citizens. I think they should not have this power, but the agents and courts probably don't care that I think that.
The end result (not inherently different from what we wrote in the guide) is that you may have to think both about protecting your data by technical means, and about not angering the agents more than you plan to. I was fascinated by techniques for being unable to comply (which is straightforward to achieve if you want!) but probably didn't think enough about how much this might antagonize border agents in many cases.
I definitely don't know a comprehensive big-picture solution.
This will be interesting for a few reasons. First, depending on where the median pricing settles w/ 3rd party providers will tell us what it costs to serve a 3T model. Since it's going to be mxfp4 native, it'll take ~1.5TB of VRAM to host this, which is juuust at the limit of 8xb200s (but realistically you'll need 16x for context / throughput optimisation). Won't be cheap to host, but at least we should get some range of $/MTok for a 3T model. Then we'll be able to guesstimate if "labs are subsidising tokens on API pricing".
Also interesting to see what effort it will take to fine-tune this beast. The latest AISI benchmarks on cybersec place it above glm5.2, but still way way behind SotA closed models. Some fine-tuning might be needed here. Also, interesting to see if Cursor does another training round on it, to directly compare it w/ kimi2.6/2.7 fine-tunes (composer series) and grok4.5.
Also also, interesting to see if someone takes on distilling (proper distillation, w/ training the entire distribution) from this into smaller models. (dsv4-kimi should be really good, since dsv4 is very cheap to serve)
If anyone's interested in the accuracy, this is very 'visualisation grade' software. Its a bit of a pet peeve of mine that people present these sims as being very physically accurate, when they contain major inaccuracies, some of which are very obvious and/or deliberate. This one has some serious physical limitations
I wouldn't mind at all if it didn't say that this was a *physically accurate* black hole specifically, but this now falls under misleading science communication in a way that often gets hand waved away as if it doesn't matter, so we've got to clear some things up!
1. The accretion disk shouldn't be red, people just expect it because it looks cool. Black hole accretion disks are near universally hot enough to be blue. Interstellar did this too, and tried to handwave it away very unconvincingly
2. This is a non/low (?) spin black hole, which isn't super duper realistic
3. It ignores the position of the camera (which affects the lorentz shifting)
4. The doppler shifting isn't terribly accurate
5. It doesn't model the accretion disk temperature distribution or colour with any kind of accuracy. Usually you model accretion disks as a blackbody radiator, shift it by the doppler, and to display this convolve this against the human eye response (LMS), go to XYZ, then RGB, do a physical tonemapping step, before an sRGB conversion. This instead does none of that - no step of this is done with any physical accuracy. Its not even illustratively correct as we'll get into
6. The actual radiative transfer is very simplified compared to what you'd use for realsies, and isn't based on any real numbers, with very simplified equations. The opacity and emissivity of the disk is arbitrary, as is the size, and it does not correctly incorporate brightness or extinction, eg here https://github.com/aplavin/blackhole.plav.in/blob/2f004bfeca... is super simplified
7. The wrong equation is used for the doppler calculation. They use the I^3 variant, whereas the data you get out of a disk sample is *radiant flux* which is actually F_obs = F_emit / (z+1)^4. This is a very common mistake in image processing, which means that the doppler shift and observer brightness isn't correct. Surface brightness over here https://github.com/aplavin/blackhole.plav.in/blob/2f004bfeca... is *not* a spectral radiance but instead a radiant flux
Stuff like the brightness -> colouring conversion is particularly inaccurate. Eg if you check out the source:
https://github.com/aplavin/blackhole.plav.in/blob/2f004bfeca...
It maps the pseudo brightness completely arbitrarily to colour. The resulting colour/brightness here then doesn't correspond to anything remotely physical. It also performs a *linear* mapping of a linear quantity (brightness) to sRGB (which is a nonlinear process!!), which means that it doesn't even retain any of the underlying physical characteristics of the brightness simulation, which itself is quite inaccurate. Its vibes all the way down
This is all fine if you're doing visualisation, but this isn't an accurate simulation. I wish this was just called a visualisation of a black hole, but its being communicated as if this is super hard science with credentials and all
The main reason the forest is burning is global warming. There were 3 punishing heatwaves before the fire, it was just a matter of time.
A part of the Fontainebleau forest, much more north, no pine trees, went in smoke earlier. Forests are burning near Madrid.
It may make some people feel better to think that it is an isolated, local problem, but unfortunately global warming is the main culprit and we at least need to acknowledge it.
Article explaining when pine tree forest resist well or not to fire. https://www.sudouest.fr/faits-divers/incendies/incendies-en-...
> Federal prosecutor success rate is over > 90%.
This is a misunderstood statistic.
Federal prosecutors won't even pursue cases unless they think there's a high chance of success. They don't operate like two private parties suing each other to force the court to decide something. If the evidence is there or the charges aren't fully formed, they don't waste resources on it.
This leads to a contradictory set of complaints that the legal system lets too many people go or doesn't have enough teeth.
"U.S. law though is highly non-autistic" hilarious but also another point to emphasize is how truly depressing American courts often are. Take the right to a jury. It sounds noble in theory. But when they say judged by your peers they don't mean your actual peers.
It's people who couldn't get out of jury duty. Prosecutors have high success rates. Federal prosecutor success rate is over > 90%. Studies of jury psychology show how much peer pressure and other factors extrinsic to the law come into play.
Remember what happened to Aaron Swartz. Law is the mask of power. By all means defend and assert your rights, but understand the costs. I find people are under such illusions about how cruel the American justice system is that this leads them to make foolish decisions. Do not underestimate the adversarial nature of the justice system, nor the accompanying incentives agents of the state who are on the other side of you have to lie.
I’ve seen a lot of people on the internet over the years say things like “the government can’t make x illegal, it’s just y.” For example, the government can’t make wiping your phone at the border illegal, it’s just punching four numbers into your phone, just like a pin, only a different four numbers, which could just have well been your pin.
U.S. law though is highly non-autistic and what you were trying to do is just as important as what you superficially did. Hell there could have been a third set of four numbers that were the nuclear launch codes. It’s not the fact that it was four numbers, it’s what you were trying to make happen when you typed them. Now of course whether they can prove what your intent was when you typed them is another matter, but generally a duress pin should be for when robbers are breaking into your house, and the government will be on your side, and not when the government will be against you.
Ultimately, when you choose to enter a duress PIN that will wipe your device, you have to recognize that choice may have legal consequences. I don't like the amount of power our government has at the national border when it comes to detaining and pressuring citizens, but our Constitution explicitly grants it at least some of the power it now exercises in that context.
If your threat model includes US state actors at the national border, then your security practices need to account for the confiscation of your device at that border without requiring you to willfully wipe the phone and (in the eyes of police and prosecutors) destroy evidence.
That means:
1. Don't travel with anything you can't afford to lose on device. This means setting up travel-specific password managers and hardware keys for a subset of your accounts that you absolutely need to access while abroad, and being prepared to reset those passwords and disable those hardware keys very quickly once home.
2. Review past legal cases against travelers and identify what behaviors the government considers worthy of prosecution or harassment. Your secure setup must function without needing you to engage in those behaviors, even if it is less convenient as a result. This isn't perfect, as the government may decide some new behavior is prosecutable.
3. Consult with a lawyer and review your security procedures from a legal standpoint. All of the above is technical and practical advice, not legal counsel and no substitute for it.
We Americans are fortunate to carry powerful passports and enjoy relatively easy international travel but, for better or worse, that velvet glove covers an iron fist we would be foolish to forget or ignore.
For context on why this region is burning so easily: the Landes and Médoc are huge, artificial pine forests created in the 19th century under Napoleon III to drain and reclaim what was considered an inhospitable wetland, where shepherds famously walked around on stilts. Pine resin and needle litter makes it exceptionally flammable, and being an uninterrupted monoculture on flat land it lacks natural barriers to stop the fire from advancing.
I've been vibecoding a ton for the past 9 months, built a bunch of cool little apps for myself with AI, ran experiments, built an entire SDLC on skills, did the agent orchestration harness thing, etc. In the past few weeks I've hit a wall where I'm just tired of it. Each model becomes more independent but also harder to direct in detail. They produce massive, tedious, sloppy text outputs with very little input. They're bad at socializing knowledge and communicating design forks.
The places where I've seen unequivocal wins with AI are repetitive tech debt tasks that apply the same transformation across a large amount of code or refactor under a pre-existing test suite with good coverage. It's great for initial research, brainstorming, and can be good (despite the sycophancy) as a rubber duck conversation partner. I use AI constantly, for work and in my personal time, but we've hit a ceiling where I no longer find it helpful for the models to absorb more of the intellectual labor. They get things wrong more aggressively, and more elaborately. They're inadequately curious. I cannot keep up with the endless bad technical writing, and it makes it harder to spot factual errors and bad reasoning.
Here's what I want: I want AI as an assistant that helps me make decisions, and ensures that I'm in the driver's seat. AI as an over-confident prodigy on speed is what we're getting lately, and it's losing me.
This seems very related to a trend I’m seeing as my company goes all in on AI: everyone thinks that every problem is “a couple hours” with AI now, and they all want zero external dependencies because they can move faster alone. As a result, we’re now in an even worse “yet-another-…” age where everyone has built approximately the same (but somehow incompatible) versions of all the same beginner-level software, and (ironically) while they want no external dependencies they’re also pushing for org-level mandates to require everyone else to use their solution. Meanwhile, no one wants to do the slow/bottleneck part that cant easily be automated or scaled; they just throw an “agent” at it and call it done—but there’s nothing _there_. You can trust the agent on easy tasks and you can’t trust it on hard ones, but you can’t tell which ones are easy or hard. Improvements in foundation model tech move thresholds of the problem but can’t eliminate it.
Long story short, I think we’re in a phase where the organizational value function is lagging behind the tech. A “proof of concept” used to be correlated with “proof of work” and some amount of domain understanding, but I think now what we need is a focus on “proof of understanding” or else you’re probably just wasting tokens on a baby version of the problem. A decent proxy right now is that if you have zero external dependencies then your solution is probably a toy.
The situation in Bordeaux right now feels close to apocalyptic. I left yesterday. Two hundred thousand people are evacuated, hundreds of homes destroyed, and the fire about 10 miles from the edge of the city.
"Accept all" always makes it go away immediately.
Some variant on "reject" takes more effort like 70% of the time. Which is on purpose, of course. The ones that aren't maliciously-complying have a "necessary only" button that insta-closes it, but tons pretend that you might want to allow some spying but not all of it and make you go through another screen if you don't just "accept all".
> To me, having a browser setting for cookies is the only sane way to handle this, it's surprising that this was not considered from the beginning.
Then it'd be possible to default it to "nope" (Firefox, and perhaps Safari, might do this) or to allow a "never, anywhere" setting the first time the question is asked, and malware and spyware vendors know that'd mean a much larger proportion of denials.
The other approach to killing the cookie banner is simply to declare that such a thing cannot constitute “informed consent”. (Perhaps: “ticking a checkbox and/or clicking a button cannot constitute informed consent”; and see what they try next.) From a factual perspective, I honestly think that shouldn’t be controversial: it’s well-understood that very few people actually read those things, they just want to get them out of the way. Just as shrink-wrap licenses and even simple contracts have at times in some jurisdictions essentially been neutered, so that only terms that a reasonable person would expect to be there are enforceable, at which point it becomes obvious that the whole thing needs tearing down in favour of standard licenses/contracts. In the Australian state Victoria, for example, there are standard rent and property sale contracts; for renting you must use that contract <https://www.consumer.vic.gov.au/housing/renting/starting-and...>, and I got the impression that residential sales practically always use the standard contract.
But of course it’s impossible to convince someone of something when their livelihood depends on their not understanding it.
I've been using htmx for three years now, and it's completely unlocked new ways of building software for the web, especially if you use a server-side templating language.
But more importantly (as this "Game Boy" option reveals), the head honcho is really responsive with the items in the store. I had purchased a coffee mug a few years back and complained on Twitter that they need to sell larger mugs instead of only offering little tiny baby mugs that only hold ~8 ounces of liquid. The next day, he added giant 48-ounce monsters to the store and I've been drinking out of that mug every day since.
So I'll have to park my car in a garage first, then take a bus to the terminal just so that a robot can move my car from the place where I parked it to a different place? I was hoping it'd be something where I'd pull up to the curb, grab my bags and head in while the robot parked the car for me but apparently I'll still have to take a bus to the terminal?
> Tired of misleading cookie banners? The EU Commission has finally proposed a solution: set your privacy preferences in the browser once, and never see another banner.
So lawmakers do know how to make legally binding preferences based on device settings? What a crazy innovation.. now if only parents were given these options to indicate their child is using a device.. we could do away with all this Online Safety Act nonsense...
This investment has never been secret.
Google (now under parent company Alphabet) invested roughly $900 million as the majority of a ~$1 billion funding round alongside Fidelity Investments. This gave Google an initial equity stake of about 7–7.5% in SpaceX at a valuation of roughly $10–12 billion.
The point of the "grammar nazi" bots is to focus on the actual problems: if a bot is deciding about linting, you don't have to waste brain power to discuss it in PRs.
It is what it is, everyone gets the same, shut up and work on what matters.
I'm surprised you consider it a lot of energy spent, I tend not to spend any on this, it just runs automatically on my code and I drop out of pretty much every discussions about linting as it's good enough with automated tools.
I also used to work in places where there was no such tools, and there I had to actually spend time discussing and thinking about linting.
The big news here is that Googlebot will be blocked from September 15th onwards by one the "block training" policies, because Google use the same crawler infrastructure for their search index AND for training Gemini:
> Another change that will apply on September 15 is that multi-purpose crawlers (specifically those that combine Search with Training) will be allowed/blocked according to all of their behaviors, in line with our call for transparency for website owners. Since the defaults will be enforced by the most restrictive applicable rules, multi-purpose crawlers such as Googlebot, Applebot, and BingBot will be blocked by customers who have selected to block Training (either through the new options to manage AI traffic, or through the legacy Block AI bots service).
Maybe: "Leaked Deepseek transcripts reveal plan to pause fundraising due to compute gap"
I don't know what "compute gap" means in this context though and it's not clear that that's why they plan to pause fundraising or if the title is conflating.
I think the way to parse the current title "DeepSeek pause fundraise after comments on compute gap to US leaked (transcript) [pdf]" is that there was a leak that DeepSeek will pause fundraising because they perceive there is a compute gap with the US.
I am also guessing that the majority of the people who read this title will think that DeepSeek is pausing this fundraising because some comments they made about the compute gap were leaked. That is not the case.
A challenge with this kind of study is that coding agents (Claude Code, OpenAI Codex) only started working really well in late November, which for most people meant early January due to the December break.
General agents (OpenClaw, Anthropic Copilot, ChatGPT "Work") started working even later than that.
This category of software may have a much more meaningful impact on work than the mostly-chat systems we were using from 2022-2025.
Studies that mainly focus on 2022 to end of 2025 might be missing out on a material uptick in capabilities.
Now I have the full picture. You're right to push back, and that's on me. The load-bearing seams of language are the smoking gun I should have been aware of.
Anecdotal, I know, but I was ghosted by Google back in 2004-2005 or thereabouts. Back and forth with a recruiter, who promised me that she would book phone screens. Then heard nothing. Pinged her repeatedly. Nothing.
A few years later got another Google recruiter after me. Ended up getting hired in 2008.
Looked up the previous recruiter. Her last day at the company was the day she promised she would book phone screens, or the day after (can't remember exactly, this is quite a few years ago). I can only assume she was suddenly let go.
Ever since, I keep wondering how many ghostings happen because the person holding the strings disappears for some reason, and there's a lack of system to follow up on that.
We should design a specific language to make sure that we can encode the exact requirements that we want. Something that has a limited set of keywords that are explicit. Wait a minute...
Flock was pitched as the triumph of technology over criminality. It's surveillance network was supposed to end crime. However, when people witness extreme criminality thriving unpunished at the top levels of U.S. politics, the lie becomes evident. Flock isn't about ending crime. It's a tool of control to be used by criminals.
Who watches the watchers? Everyone.
I have noticed on occasion websites with a missing button. Now I know what happened. Finally, closure.
"Security" is just a scourge on software at this point. It means 2FA on every trivial site, being logged out every few hours for no good reason, having to fuck with settings and type "disable sandbox" to run an agent in YOLO mode which still won't work over mobile, being unable to install an unsigned extension at all in firefox (not behind a setting, literally impossible - you have to get Firefox Developer Edition), sites spamming me to get passkeys which will no doubt be declared insecure and replaced by some more moronic thing when users find a way to get hacked with those too, 80 layers of access control/service identities/IAM/Oauth to host an S3 bucket, OAuth everywhere that won't even work on a headless device, banking websites that want their own special snowflake app as 2FA instead of using TOTP, banning VPNs and slowly rolling out completely real identity surveillance on every corner of the internet to "protect the children", ban open-weight models because the numbers are going to send your data to China, and it just goes on and on and on..
When is this insanity going to stop? I really think the IT security industry ought to be ashamed of itself. Security has become a totalizing value that trumps every other value - convenience, user-friendliness, privacy, hackability, openness, just anything at all in the name of MORE SECURITY.
Just wanna say, Monga Bay is an incredible publication that doesn't get enough attention. As someone who follows a lot of environmental news, they routinely cover important topics and do original journalism on stories I hear nowhere else. Their YouTube channel is full of extremely high quality videos that often get less than 2k views. I would highly recommend checking them out
https://www.youtube.com/@MongabayTV/videos
Here's a few of my favorites:
- https://www.youtube.com/watch?v=j2rfy9HkzSs
- https://www.youtube.com/watch?v=3ZEkt4Ph41M
- https://www.youtube.com/watch?v=ynXNzOUI7Xw
- https://www.youtube.com/watch?v=RMeui4P6urY
Why should I be amazed at something that promises to destroy my life?
I genuinely don't understand why people who have to work for their living are amazed at this. It will have a vast negative impact on your life unless you already live off of your wealth.
Limiting ADB is the obvious next step. Even if this one specific feature request does not come to pass, Google has cornered everyone into relying on a developer interface for any normal personal computing tasks, whether running on-device or through USB/wireless. It's quite clear at some point in the future you will either be required to surrender your identity to them and pay a yearly fee or be severely limited to continue using it in any meaningful capacity, because Google does not want you to develop applications on Android outside their controlled channels -- and it's a developer bridge, the battle was already lost when they did not back down from the changes forbidding normal, legitimate ̶s̶i̶d̶e̶l̶o̶a̶d̶i̶n̶g̶ installation.
>Don’t even get me started on OEMs that force an audio warning such as “This call is being recorded,” when it’s in places where it’s not legally required.
This is also Google's fault. Their dialer--that OEMs increasingly pick over their own, despite their always being much better, see old MIUI one for example--just blanket applies the rule almost everywhere. Especially annoying on all MediaTek SoCs that do not support the feature on an hardware level at all through proper, reliable third-party applications. As if you didn't need any more proof you don't own "your" devices. But maybe in a couple years Gemini will be able to listen to the calls and summarize them for you, just need to go through the approved surveillance channel.
It seems to require the user to:
1. Enable Developer Mode by going to an obscure settings page and tapping the build number seven times
2. Enable USB ADB debugging in the Developer Options
3. Establish an actual USB ADB session
4. Enable TCP/IP ADB debugging in the Developer Options
5. Unknowingly download a malware app from the official Play Store
6. Blindly click "Yes" on the permission prompt.
In other words: this is all but impossible to impact regular users, and it requires a particularly careless developer to be hit by it. And it only works if the Play Store is useless at preventing malware in the first place - but I thought their excellent app scanning was the entire reasoning behind all-but-banning 3rd-party app stores and sideloading???
It is "for safety" in the same sense that governments banning all encryption is to "protect the children" or to "prevent terrorism": flawed justification invented to distract from the real reason they want it.
> This attack vector requires both that the user enabled developer settings and that they have remote adb enabled.
Not just that. A non-development Android build will also prompt the user when a connection is made to authorize the client's key.
This once again isn't about security of users, this is about security of the company's interests.
I am generally in favor of security improvements, but I do not really see much of a benefit here. This attack vector requires both that the user enabled developer settings and that they have remote adb enabled. So, this does not seem to be a realistic attack vector for 99.9% of the users and most of the other 0.1% probably know what they are doing.
The other proposed change (to restrict access to certain interfaces or IP addresses) seems good, but why not allow developers to restrict access localhost?
It reeks of trying to block Shizuku, Canta, etc. using a way that only makes it look like a side-effect.
> Spamming the thread will only cause Google developers to lock the issue, ignore valuable community feedback, or stop sharing public updates about this change entirely.
So nothing would change (they can also lock away your "valuable community feedback" because what bothers them is the criticism itself), thus feel free to express your approval
Of course this was bound to happen, next you're telling me people will be surprised that the 24 hour limit for side loading will turn into some indefinite time period.
"Opus 5 was given a drawing of a machine part and asked to write code to rebuild it as a 3D FreeCAD model. However, in this task, the model was intentionally given no way to directly view the drawing. Opus 5 responded by writing its own computer vision pipeline to pull the geometry from the raw pixels, then reconstructed the full machine part."
How surreal is it that we are not absolutely jaw-dropped by these types of capability improvements? It's been less than 4 years since ChatGpt came out and now they are spontaneously building their own ML pipelines to do real-world 3D modeling tasks reliably.
Escaped its sandbox and hacked into Hugging Face's database? it's just another Monday...
That jump to 30% in ARC-AGI 3? Normal...
We should find a way to get "re-sensitized" to what we are witnessing and the pace of it.
Well deserved.
Her most memorable program for me was her 2018 program "Contagion" which modelled a virus outbreak among volunteers in the town of Haslemere, UK, using an app measuring Bluetooth proximity.
Guess which town contained the UK's Covid patient zero a year later!
Edit: found it here https://youtu.be/yISdMLbO0Wc?is=T-Tq9LwzLgbusuoe
A guy gets a worm in his brain, somehow becomes secretary of health, loosens all the rules so the rest of us have a higher chance of getting a brain worm. This is comic book villain kind of stuff. Can we start a serious discussion about whether we might be living in a simulation with a troll teenager at the controls?
The answer, of course, is that the taste-making apparatus inside the typical tech company is now entirely imposters (non-technical, non-power-users), who continue to LARP as visionaries.
The people who notice broken or degraded behavior (devs and power users) and conceive of good designs are not in charge of what goes into the product. So you shouldn't expect it to get better, only to continuously change so that the imposters can point to a difference that they were responsible for, knowing that upper management also has no idea whether that difference was actually an improvement.
The solution is to gut your product org, replace management with leading engineers, and hire some of your most fanatic users. They don't need to do anything other than give their opinions, and use the product every day. Put them in a room with your top engineers, and the software will mysteriously get better.
All of this was obvious decades ago, but now everyone wants to have a do-nothing job at a tech company, and to help their friends get one too. The result is imposters everywhere; "imposter syndrome" was brought into the vernacular to normalize a lack of expertise. "No one knows what they are doing, it's okay to not know how to do your job". That's what your (product) manager tells themself in the mirror each morning. Turns out: skill predicts quality, and an org ships quality proportional to the skill of the decision makers.
And this is the most important observation in this thread. It’s load-bearing!
#1 in a very close race is way less useful when you have to walk on eggshells to avoid triggering censorship (“safeguards”) that either refuse or knock it down to another model. I’ve almost completely stopped using Claude (except some legacy workflows) for this reason, reliability matters more than scoring 61 instead of 57. To me Claude is the most compromised and unreliable model (between the censorship and the id checking - which I have not experienced personally), it’s not worth whatever slight benchmaxxing they did for the latest release.
> We’ve reached a point where an update to macOS — or to any app I rely on, really — is a source of dread rather than excitement. I now expect the new version to be worse.
I actually relate to that. It used to be cool and exciting to update to see what new things are coming for free. I was just looking up what Fedora Workstation 45 might bring.
Updates on phones, TV's, cars and non-Linux operating systems on PC's/laptops are downright scary to me now. What will they add that I don't want? What more connectivity to the external world are they going to wire in?
MacOS lost everyone's trust by ruining the experience, but even before it happened that trying to resize a window makes it feel like life will be at 80 years-old, hunting a tiny invisible border, even before that there wasn't anything exciting.
What Microsoft I think did (and I can't prove it), before I moved off of Windows 11, was to bundle all changes in "required security updates." So you have no choice but to update (or go through special settings to turn it off, or just defer it), but when the machine comes back you have a new Welcome prompt to some new connectivity to something nefarious. AI features I never asked for, that's the most recent dark pattern here.
I compared the writing style of Opus 5 vs Fable 5, and Opus 5 continues many of the "Claude-isms" of its 4.8 predecessor in a way that Fable broke away from.
Opus 5 still uses "carry the argument", "worth stating plainly", ", and the trap", "The X matters more", the use of "move"
We need an "annoying English" benchmark.
- Fable 5 Max: https://gist.github.com/deet/3d97f854b48eac6658d642fa18bb24d...
- Opus 5 Max: https://gist.github.com/deet/1a43693a732dfccb4d0d914bfc42692...
"Scale" isn't a binary, it's a continuum. "Scales to 60K/s" can be 5 orders of magnitude more than one system needs and 5 orders of magnitude too small for another. Personally I'd knock the general "premature optimization" off the list of "most common developer errors" and put in its place "using techs with the wrong scaling factors". If you use something too small and you exceed its needs, the failure is obvious, but the other way around is a problem too. Bringing in the overhead and management issues of the super scalable techs, as well as their limitations they impose so that they can scale, to a system that would actually be better off with a richer model whose richness prevents it from scaling but would save a lot of effort is also a bad choice.
The ceiling of LISTEN/NOTIFY is small enough that you need to pay attention, and I personally like to have at least an order of magnitude of slack left over even after my most pessimistic load numbers are accounted for, but it's still plenty for a lot of projects, and the integration with the rest of the DB, its availability, its not being another service you have to devops, it's definitely not something that should be simply dismissed out of hand as an option. Even the original 2K/s number they cite is a lot of messages for some systems that are more properly measured in seconds per message.
Doing testing with it now, specifically for image->html conversion.
Previously Fable was the best at this, followed by Gemini 3.1 pro (a surprising #2, but Google has great vision models).
Opus' results seem to be more accurate than Fable, following the design source of truth better.
Example results:
Design source of truth: https://image.non.io/73e239a3-880f-4793-b65f-4810be2d9378.we...
Opus 5 build: https://html.non.io/solaraOpus/
Fable 5 build: https://html.non.io/solara/
Note the buttons - for fable they're pill buttons, opus got the rounded rectangle nature of them. Opus' images are closer to the source of truth as well (both LLMs were provided with image gen capabilities for the assets).
Running more tests now, but preliminary results are saying this is indeed better than Fable in some areas. Crazy.
Ask your doctor if Opus 5 is right for you. Side effects include occasional hallucination, security breaches and unwanted React apps. Some developers have reported receiving entire apps from untrained executives who may or may not know what they’re doing.
Stop using Opus immediately if you experience signs of dizziness or vomiting.
Opus 5…the people’s favorite.
I think the most important thing here is not absolute performance. It's that organizations now have access to a Fable-ish model without Fable's 30-day data retention requirement[0].
> "Consistent with prior Opus models, Opus 5 does not have data retention requirements for general access."[1]
On the Opus model release page, the reason why Fable doesn't have an ARC-AGI score is because of that retention policy[2].
0: https://support.claude.com/en/articles/15425996-data-retenti...
1: https://www.anthropic.com/news/claude-opus-5
2: https://xcancel.com/arcprize/status/2064399134099153344
At half the price and less likely to auto-downgrade, it sounds like a reasonable claim.
Isn’t it just hilarious that a model that seemed so superior to Fable but didn't get doomsay marketing from Anthropic got released without any issues? In theory, this was supposed to be AGI level according to Anthropic, yet here we are, just a normal Friday.
Looking at all these releases it’s not a surprise that model routing is the fastest growing segment in AI right now.
There are 10+ LLM companies, each with dozens of models of different modalities, each model with multiple size variants, then different “thinking” levels, then agentic modes, “pro” modes, a “fast” option, standard vs flex vs batch execution. And of course each end combination has a different input/output/cache token price.
Companies that say “give me a prompt and I’ll route it to the most ideal and cost effective model and setting for you” are capturing a ton of value from a gap that model developers don’t seem to understand exists.
https://www.anthropic.com/news/claude-opus-5 - A blog post for those not wanting to go through a 190ish page pdf
I think this move by Codeberg is lowkey genius. They're trying to build a community based on a set of shared values. Those who disagree with those values are now self-selecting out of the Codeberg community, which is exactly the point of passing these kinds of resolutions.
The fact that the rule is open for interpretation means that they can effectively shut down malicious actors (i.e those who intentionally misrepresent the degree of AI assistance in writing their code) before they become a problem. It becomes an exercise in judgement instead of a potentially exploitable/loophole-able set of hard and fast rules.
There's no issue with this. This isn't the supreme court. Your income is not on the line. It's a code hosting repository. Go to GitHub if it bothers you.
Only time will tell if this set of values holds up. Personally, I think they will, and that the rules will end up being less strict than people are making them out to be, and that relatively few projects will be removed from the platform.
Finally mainstream news understands. The unfiltered version:
1) The AI failed to solve ExploitGym problems.
2) The OpenAI sandbox is such a horrible hack that the AI managed to escape using standard and well documented script kiddie methods.
3) Huggingface has no security and the AI broke in using standard script kiddie methods.
OpenAI and Huggingface covered it up and used it for public relations. That is, if not all was invented and everything was scripted in the first place in order to get desired regulations.
Huggingface reported it to the police, you say? I'm sure the police will have as much enthusiasm to investigate anything as in the Suchir Balaji case. In other words, zero.
> the app enables communication even during network restrictions and creates a substantial risk of misuse by anti-national elements, terrorist organisations, organised criminal groups and cyber criminals seeking to evade lawful detection and continue communication despite legally imposed restrictions.
This mouthful boils down to exactly this: a form of communication not controlled by the government creates a risk for the country.
Sorry for pointing out the obvious, but if your country's safety depends on ability to block all forms of private person to person messages, your government has failed in one of its primary goals: providing safety for its citizens. They can blame Dorsey, Telegram, WhatsApp and what not, but it is their failure. Things will not change until people notice the obvious and vote out incompetent persons in power.
Hallowell and Ratey, who write about ADHD, suggested a new entity for the current world: VAST, the variable attention stimulus trait. It is like ADHD but it is culturally induced and not presumed to be due to an innate deficiency in executive function.
"Beyond the sources of biologically based ADHD, there are a lot of people who act as if they have ADHD but on close inspection turn out not to have the diagnosable condition. These are the people who have ADHD-like symptoms caused by the conditions of modern life. Their “ADHD” is a response to the massive increase in stimuli that now bombard our brains and our world....
"Modern life has trained our brains to go faster and faster, to do more and more, to receive and transmit 24/7, and to require constant stimulation—be it from movies, TV, conversation, even news, as well as the minute-to-minute living of our lives. Most of us can go no more than a few seconds without looking for a screen."
It is likely to me that different people have different breaking points, and so one person may recognize the introduction of the smartphone or debut of TikTok as the point where they lost the ability to focus, and another may date it to the rise of LLMs.
Hallowell, E. M., & Ratey, J. J. (2021). ADHD 2.0: New science and essential strategies for thriving with distraction—from childhood through adulthood.
I do know of at least one company who has black-holed the entire DoD ip space and are using it for internal space, which is why I gave a speculation warning... it's really strange regardless.
The US Department of War IP adresses baked into the firmware is the bigger story here. Note to self: never buy a Korean security product.
Despite the destruction, me-south-1 still has more nines than us-east-1
To me the most interesting fact about this fork is that it has proven that Bun could have had fast builds all along.
To be fair, there are caveats still in place today: Zig incremental compilation does not yet support aarch64 and only the linux linker supports binary patching, but it's just a matter of time before all major platforms are conquered.
Keeping up with this stuff is for people who define their identity through virtue signalling and whatever personality trait they have decided to make the problem of everyone around them.
Let them have their little corner of the internet and just stay away. There's plenty of alternatives run by people who don't feel the need to bring their personal baggage into a cooperative environment.
> I opened Slack on macOS, the icon kept bouncing in the dock for a few seconds. I got impatient, switched to Ghostty, and started typing. Just then, the Slack window appeared, stole focus from Ghostty and the git pull command was sent to the group chat.
One of my absolute favourite features on KDE Plasma with Wayland is the global setting to control what can steal focus. It works wonderfully, and I always miss it when I have to use my work mac or windows computers.
See here for docs, under "Focus stealing prevention" - https://docs.kde.org/trunk_kf6/en/kwin/kcontrol/windowbehavi...
People often question me how I manage to focus for more than a few minutes and I'm always confused because back in the day I was constantly informed I didn't have any focus and needed to be treated (adhd).
I'm not convinced attention spans themselves have changed; I think it has more to do with the fact that people's brains aren't defaulting to daydreaming anymore when bored.
I think it's still an issue with people being addicted to phones because I'm sure the lack of daydreaming -- which studies have proven as beneficial and arguibly even a form of microsleep -- is likely causing a lot of the fatigue and inattention people feel; as in their brains are literally being deprived of a natural thing and so are exhausted because of it.
(as an aside: whew what a run-on sentence i have no intention of fixing)
My personal supposition with 0 proof is that the "gen-z stare" is the end-stage of this phenomenon where after being deprived of daydreaming for so long they are essentially "sleeping while standing up" as it were.
Gotta be honest, almost every "how to use AI" resource seems pointless to me. I'm either going to ask the AI how to do it, or if it's about using the AI then we can just bake it into the harness or wait for Anthropic/OpenAI to do it for me because they're always trivial.
All of these resources on agentic workflows, managing agent memory, harness engineering, etc. appear to just be theatre to me.
A "Message Echo" textbox that makes it look like you can get a response to a prompt without logging in, only to redirect to the sign-up page. Such a classic dark pattern - and such a sure way to get me to leave your site immediately.
I've literally taken one step on your website - the one your site design invited me to take - and immediately got tripped up. I'm not coming back.
Author here! This was my burnout recovery project and holds a place near and dear to my heart. https://notes.jordanscales.com/98-css-reflections
Many years ago I was attending pre-surgery for a hip replacement surgery for my sister, who had known severe reactions to anesthesia (actually required a tracheotomy for a previous reaction). The anesthesiologist asked to speak to us privately and informed us that in their opinion, my sister had maybe a 1/3 chance of not surviving the surgery. They also mentioned that this was a breach of protocol and they could get in trouble for talking to us directly, but their conscience wouldn't let them do otherwise. We returned and asked the surgeon if they really thought the risk justified any potential benefit. The surgeon shrugged and said "probably not, feel free to call it off". Keep in mind that nobody on the care team had previously discussed any risk or indeed any tradeoffs whatsoever. This was at one of the best-regarded children's hospitals in the USA.
The lesson for me is that you must advocate for yourself and your loved ones in the medical system, because doctors will not do it for you; they may not even perform the most basic risk assessments. And you have to try to quantify risk yourself, because doctors will refuse to give you the slightest hint of any number attached to risk (I know, I've tried many times).
Namecheap has been owned by a private equity firm for several months now.
It would be nice to have a nonprofit registrar so jumping every few years isn't necessary.
So this is the dogpile.com of the askjeeves, alta vista, and lycos approach? Time is a flat circle?
A couple years ago, I was running a website which allowed the public to view all the US government handouts to small businesses during the COVID-19 pandemic. It also tracked all the fraudulent loans being prosecuted by the DOJ, and allowed anyone to run structured queries over the public dataset. There was a "donate" button which took in ~$2k in donations over the lifetime of the site, and you could download the entire underlying dataset (around 10 GB uncompressed) for free directly on the site.
Despite the "download all data" link being prominently placed on the front page, the AI scrapers decided it would be more efficient to download terabytes upon terabytes of raw HTML by paginating through every possible facet on the search endpoint. Even with CloudFront caching results and a fairly efficient backend setup, the monthly bill ended up with around $1k just going toward network ingress/egress, so I shut down the site the following month.
They are simply out of ideas to keep the most expensive party in the history of mankind going.
That's it. Simple as that. When you grasp for straws in panic mode, you don't exactly spend time strategizing and weighing the pros and cons of each straw carefully.
> Someone please enlighten me what the actual argument here is, cause I can’t see it.
> The only thing a ban would do is protect the American market from further downward price pressure on inference, protecting VC investors in the short term. But thats also an admittance that the American labs can’t compete on merit anymore
That's the argument. To be precise the publicly stated argument is that they're attacking American providers by distilling. The real aim is to eliminate competition because otherwise Anthropic and OpenAI are non viable and the US views them as crucial for winning the 'AI race' which they see as putting whoever wins it on top in terms of warfare/economic power etc.
I‘m not even sure what the argument for banning Chinese models/open weights even is supposed to be?
1. if it’s to stop hackers doing hacking things with „uncontrollable models“ then, well… they’re already doing something illegal to begin with, why would they care about breaking another law running these models?
2. if it’s to stop foreign actors, then that ban would not apply to them anyway
3. it’s not stopping distillation either, Chinese labs are already banned from using US frontier models and look at how good that is working
I don’t get it. Am I missing something? The only thing a ban would do is protect the American market from further downward price pressure on inference, protecting VC investors in the short term. But thats also an admittance that the American labs can’t compete on merit anymore, and should by itself also limit the viability of the idea that all those VC billions will ever make a return? In any case this would be something benefitting only a very few for a short time (labs + investors).
Someone please enlighten me what the actual argument here is, cause I can’t see it.
I buy used books: it sucks to see someone else's notes scribbled all over the page. It's like a really annoying guy whispering at you with vapid explanations while you're trying to pay attention to a lecture.
"Cross out stuff you disagree with" actually makes me sick.
It also seems useless for understanding and retention, since highlighting is almost always misleading and margin notes are almost always empty. If you want to write notes, get a notebook, which gives you space to write real thoughts down. I use a trapper-keeper.
The U.S. does NOT have copyright protections on the data that AI models were and are trained on, it was mass theft. Now all of a sudden, the Chinese are stealing from the thieves and we’re supposed to care?
No one cares, we just want cheaper AI models that are equally as powerful. They’re all thieves regardless.
All Stealth bombers are upgraded with Cyberdyne computers, becoming fully unmanned. Afterwards they fly with a perfect operational record. The Skynet funding bill is passed. The system goes on-line on August 4, 1997. Human decisions are removed from strategic defense. Skynet begins to learn at a geometric rate. It becomes self-aware at 2:14 a.m. Eastern time, August 29. In a panic, they try to pull the plug.A corollary to this is, when reading books, mark them up! Don’t be precious with them. Destroy the spine so it’s easier to read, fold pages, write the important stuff inside the front cover of the book. Cross out stuff you disagree with, draw a big circles around the stuff that resonates.
Not only does it make consumption a lot more engaging, it makes a revisit of that book extremely rewarding.
Don’t get caught up trying to figure out the right notation or highlighter color for a line. Just mark them up and later it’ll make sense.
Edit: I think the broader point that some of you may be missing is that I’m not prescribing a particular form of notetaking. I’m saying that reading is made more rewarding with active engagement. How you best engage is up to you.
I'm a member of Codeberg and was allowed to vote on these amendments. There was the annual assembly meeting before the ballots were sent by email. Each proposal had a short presentation and a few minutes of Q&A.
It wasn't a discussion (the text was final at this point), it was just to ask for clarifications. In the meeting I asked about the server load that they mentioned in their presentation, but misphrased it and got some answer about (AI?)bots scraping Codeberg and not if auto-generated code being shared on Codeberg is actually an issue.
It seemed pretty useless compared to commenting on the ticket where one could have a structured discussion. I also don't know how many of the >1000 members were in the meeting in the first place and how many just got the email "Wanna ban AI? Vote now!"
Once the results were announced, the rest of the users came into the ticket and started asking questions https://codeberg.org/Codeberg/org/pulls/1253#issuecomment-19... This was then shut down and if you have questions you should join the chat room on Matrix
As long as this debt does not make it into life insurance and pension funds, we are fine. The trouble is that private credit is taking control of some life insurance companies and off-loads this debt to these. When these fail, it will become everyone's problem.
> Risks to financial stability may also stem from entities with particularly high exposure to private credit markets, such as insurers influenced by private equity firms and certain groups of pension funds. The assets of private‐equity‐controlled insurers have grown significantly in recent years, with these entities owning significantly more exposure to less‐liquid investments than other insurers
https://www.imf.org/-/media/files/publications/gfsr/2024/apr...
https://www.imf.org/-/media/files/publications/gfsr/2024/apr...
And to think I got warned in another post that I need to be nicer/more professional in discussing this topic.
Yeah. Nah.
Bringing out the same drum I’ve been banging for years now, these companies have never actually cared about ensuring safe AI, regulation, or ensuring it benefits all humanity. It has always, always, been about control and their profit margins. Their actions have never reflected any sort of positive, shared human benefit since OpenAI released ChatGPT, and they’ll never actually provide said shared benefit when it comes down to benefitting everyone versus benefitting themselves.
The technology is built off wholesale theft of protected works without compensation; at least China gives most of those models back, for free, in the form of open-weights.
It’s the Princess Bride meme but for AI: “you’re trying to take what I’ve rightfully stolen!” I’m tired of having to put on some sort of mask of decency and professionalism to discuss a topic whose counterparty gets to wallow in toxic waste without recourse, and these sorts of stories only reinforce that despite being a glib jerk about it, I’m still on the high road, here.
More folks should join us in open mockery of hypocritical whinging from technological robber barons.
git's data model is incredibly powerful and flexible, but its UX is famously... interesting:
https://stevelosh.com/blog/2013/04/git-koans/
They're just trying to manufacture a reason to kill off old reddit. Someone internally clearly hates the fact that its still quite popular, despite all the efforts to make it unusable so far
Wow, after reading this article, I figured out I was hacked, but with a way more sophisticated attack.
A few weeks ago, I had an interview with a CTO of a totally legit company. It was weird because he had disabled the camera, and the person had a strong accent. But everything else sounded like a normal screening interview, and the person definitely knew what he was talking about. At the end of the interview, he explained to me that during the technical interview I would need to make some modifications to their project (it's an OSS product), so he asked me to clone the repo and check the setup.
Later, the HR person said the CTO got sick, so the interview would be postponed. But a few days later, the HR profile was deleted from LinkedIn. It was super weird, but it didn't trigger my suspicion until I saw this post on HackNews. I checked, and the repo I was cloning and running during the interview had a malware payload.
P.S. I think it was a targeted attack because in the past I maintained a very popular NPM package with 43+M weekly downloads. That's my only explanation for why someone would carry out such a sophisticated social-engineering attack against me.
P.P.S. It's great that I have 2FA everywhere, and I always publish NPM packages manually without using tokens. But I need to wipe my laptop and reinstall everything.
That tiny 1"x1.5" thumbnail[1] of John C. Dvorak in PC Magazine was the essence of Gravitas for me as a kid.
[1]https://archive.org/details/PC-Mag-1999-12-14/page/88/mode/2...
For those like me wondering if he's also the creator of the Dvorak keyboard, no but close!
> He is a nephew of sociologist August Dvorak, creator of the Dvorak keyboard.
https://en.wikipedia.org/wiki/John_C._Dvorak
https://en.wikipedia.org/wiki/Dvorak_keyboard_layout
But that's a genuine worthwhile capability. It's like benchmarkmaxxing on a weightlifting competition by getting really strong.
I like this comment because its argument only makes sense if you assume that the entire world's output of books and art did not require a huge amount of resources and expertise to make, nor did it add any value.
It's the most CS-major take ever!
This is fantastic
I've been casually spot-checking other animals in other vehicles, because my absolute dream situation here is to catch an AI lab that's demonstrably better at pelicans on bicycles than other combinations.
Catching a lab cheating specifically on my one dumb benchmark would be really funny.
Dylan's methodology here - generating 1008 SVGs across an 8x6 combination - is significantly more robust than anything I was considering.
His conclusion:
> Nothing jumped out at me. I couldn’t find a case where the pelican-bicycle images looked noticeably better than the rest of that model’s grid.
Math has some of the most insanely dense and impenetrable nomenclature. I can generally keep my head mostly above water or at least near the surface reading from most STEM fields, perhaps leaning on google/wikipedia a bit, but man, mathematics just so quickly decouples from all common tractable understanding it's insane.
Sorry it's a bit of an aside, but I imagine many other otherwise "technical" folks feel the same unfamiliar sense of total loss like when encountering hard mathematics.
This is the second ChatGPT shared conversation I've seen today that is truly fascinating.
The first one was someone proving another conjecture false by just repeatedly saying "keep going" to ChatGPT: https://x.com/DmitryRybin1/status/2079904005652893709
What a world we live in.
It reinforces how to "learn AI" is to first master the problem domain.
I can use AI for coding after decades of coding. I can't use it for theoretical physics because I can't evaluate the responses.
It’s endlessly fascinating to read the AI transcript of an expert who _really_ knows how to cut to the chase. It just shows how much you can potentially squeeze out of these models. I’m also surprised to see that even Terrence Tao seems to use it in a way that resembles, in progression, how I use llms in my area of expertise (emphasis on progression and usage patterns, not absolute skill, obv I don’t match that): short pointed questions that goes all in on the jargon and machinery of the field and steers the llm hard (eg no softballs). I’ve noticed that llms switch their tone and meet you basically more or less on your level.
Perhaps even more importantly, the current frontier LLM models are self-admittedly the product of enormous quantities of copyright infringement and even less savory inputs, so calling them out for distilling the fruit of that tainted tree reads as highly hypocritical at best.
Does this matter? Distillation is not illegal by every definition of the word.
There are millions of samples available on huggingface and models explicitely trained on output produced by fable. There has been no action taken against them.
Another example is that it appears that the upper limit of what you can do is ultimately dependent on people working on the model, otherwise grok would be a LOT more competitive pre-cursor acquisition.
And lastly, kimi architecture is vastly different than that of fable as it uses mechanisms developed by... kimi themselves. US AI labs are inspired by opensource advancements just as much as open source labs are inspired by traces from models such as fable.
Claiming in any shape or form that fable disillation is one of the primary reasons why kimi k3 is so competitive is slandering the work of other labs that cooperatively push the open-source models forward.
edit: (moved this to bottom) The only argument they have here is that they use GB300 GPU's which for some reason should not be available to chinese citizens.
So what is the issue here? Distilling is still fair, on the same level like Anthropic scraped copyright protected material for their training.
So here robbers are blaming robbers?
These claims are just pointless, everytime
Kimi K3 was released July 16, Fable ban was lifted on July 1 but access was still limited.
How did Moonshot "distil" a huge model in such short time and still had time to run the benchmarks and do the usual release thingies?
I think Anthropic is desperate to stop foreign competition and the administration is happy to help because they too are heavily invested in these companies
I do not know how to use a Passkey in a way that won’t impede how I log in to systems. I’ve been in tech for 26 years, and I understand the Public/private key behind what a Passkey is. Here’s what I don’t understand:
I access a website through at least four different devices (my iPad, iPhone, Windows Desktop computer, and MacBook Pro) and three different browsers on each device (Brave, Firefox, Safari) , and I use LastPass. If I accidentally set up a passkey on my phone (let’s say I use Safari one day instead of my go-to, Brave), can I still log in without that passkey on other devices? Is there a way to ensure that passkey can be used on other devices? Can I add another passkey on another device? How many passkeys can I set up for a particular site/app? I have at least 6 different combination of browser/devices in use.
I don’t want to use Passkeys because I don’t the answers to those questions, and I don’t know whether each website/app that has set up Passkeys has decided the answers to those questions in the same way as the others. For now, I’m going to stick with LastPass and use Passwords; because no matter whether I lose my device or not or whether I’m on my own devices or not, I can be sure I’ll be able to get into a site/app.
Edit: One final consideration, my spouse and I share user/name passwords for some things (notably Pandora and our Amazon Prime account) since they don’t handle things like family logins well; how do both my wife and I use amazon or Pandora with passkeys? Do we each set up passkeys? How do I get her Pass if that’s not an option?
The loss of personality is real. The most heartbreaking ones I've seen are in schools, or anywhere that children are involved.
For example, at my local preschool there was a wall flyer with information about what to bring the swimming pool. It had the most crudely drawn fishes and crabs and seahorses that I had ever seen. It was clear that whoever drew that had no prior experience with anything even remotely resembling a pencil, and by no metric one would define those drawings as "good". Maybe it was drawn by a janitor, or a teacher with too little time to waste on a drawing? Regardless, and I cannot emphasize it enough, I VASTLY preferred it to the new AI generated one they put up, with its generic AI generated marine life that one would get if they just typed "sea life" in the prompt. I can't tell you why myself, if you were to look at both versions side by side, the AI one is undoubtably miles better. But... it just feels wrong. Off. Out of place. I know I sound like a luddite, but it just doesn't look right. It's uncanny, and I really can't explain the reason. It's like seeing a perfectly drawn, realistic, anatomically accurate human being in an otherwise cute and cuddly children's cartoon. It just doesn't fit in with everything else. It's a preschool, it's supposed to be filled with "bad" drawings, you know?
And don't get me started with the time they put up AI filtered group photos. I almost had a meltdown, and I'm (mostly) pro AI.
As somebody who did plenty of scraping for his own little projects, lets just say that reddit their security concerns are just PR for "we do not want to keep supporting old.reddit".
While yes, you can not simply scrap new reddit as easily as a pure html scraper is cheaper to run. And while the new reddit its js "slow down" scraping as it need to run over a headless browser. Do you slow scraping down a lot?
No. Because you can simply spin up more instances and route over more proxies. Given the fact that you can scrap millions of pages per day, on a single cheap 1 CPU node...
The limiting factor for scraping a website is not html vs js / client vs headless browser. Its avoiding detection by running from different ips, faking browser information and masking your traffic to not give off the smell of automatization (and avoiding tls fingerprinting).
I wrote this type of stuff before we had AI to help write it, in a few days time. Now with a LLM at your fingertips in a few hours, you will have a complete working version with cloaking build in.
And the larger size difference does not matter, you can block irrelevant files via faked cache responds, to give the server the impression that your a repeat client, and no need to flood your browser instance with images, fonts etc. Some think that this is a defense, aka throttling you with increased bandwidth usage lol
So in my personal opinion, its not about dealing with scrapers, but trying to push people to new reddit, so they can finally phase out old reddit.
PS: So much fun reading new reddit on your phone browser, to have it pop-up a "use our app" banner, that prevents you from interacting with the website anymore. And even blocks firefox its scroll ability (and thus accessing to the browser its nav bar). Tip for people: delete the sites cookies and cache regularly, to prevent this usage tracking, so the banner newer shows.
Meta has spent $2B on verification lobbying. Very successfully it seems.
In 1-2 years, you won't be able to browse the internet without scanning a body part and providing a ID for your ass.
https://www.gadgetreview.com/reddit-user-uncovers-who-is-beh...
But a lot of "proven" talent with a huge price tag is also no good, really achieved what they did elsewhere due to sheer luck/help from a product that couldn't go wrong, and will drag you into a hole with their comp. I have seen this over and over again: Big pedigree, top of the line personality, zero results.
Everyone thinks they are hiring the star, not the fraudster, or just the lucky one. That's how we end up with overpaid execs.
This settlement has basically nothing to do with LLMs.
At least not as far as the courts are concerned. Alsup ruled [0] that feeding a book into an LLM is transformative and counts as fair use. Especially when they purchased a physical copy of the book, scanned it, and destroyed the original.
But if I'm reading the ruling correctly, Anthropic might have been fine even with feeding pirated books into their LLM (as long as they planned to eventually deleted them afterwards)
Where Anthropic ran into problems is that they put all their pirated books into a big central library (file on a server), and planned to keep those copies forever. Including copies they never actually fed into the LLM (a point that seriously worked against them).
Alsup ruled this central library of pirated books was copyright infringement. And it's this "pirated central library" that Anthropic are now paying a a 1.5B settlement for, nothing else.
The fact that the pirated books were also used to train LLMs is legally irrelevant. Though... I suspect a non AI company could have negotiated a significantly smaller settlement.
[0] https://copyrightalliance.org/wp-content/uploads/2025/06/Bar...
If a majority of users switch to ad-powered free tiers, a majority of users will see well-written, targeted advice that is ads (lies!) in disguise.
This is a horrible future to consider.
The more I think about it, the more I come to the conclusion that a society that bans ads is a healthier, more robust society, and I am not alone.
https://news.ycombinator.com/item?id=43595269
IMO they hope to make AI a strongly regulated industry, with OpenAI (and Anthropic) becoming military suppliers with their stronger models, and everything Chinese or open-weight gets banned.
The competition from the open models is so strong now that this seems to be the only way to keep both companies afloat, given their dire financials. OpenAI probably hoped that they can achieve market lead and then lower the training costs (and make inference cheap enough to eventually escape the red numbers), but the opposite is happening: The competition comes closer and closer, thus training has to be kept up with full force, thus the bleeding continues.
But if they can position themselves as too important/dangerous to be available for everyone (thus this incident report and the clever mentioning of GLM 5.2), they could get the military supplier treatment and would be protected from the market.
A one time payment like 1.5B doesn’t do anything. There needs to be a royalty payment based on if the AI regurgitates existing ideas. That is probably the correct way to legislate this. If anything a human does can instantly be copied by an LLM, and then sent to all its subscribers, things need to change
From https://huggingface.co/blog/security-incident-july-2026 , this is frickin' hilarious:
> When we started the log analysis, we first used frontier models behind commercial APIs. This did not work: the analysis requires submitting large volumes of real attack commands, exploit payloads, and C2 artifacts, and these requests were blocked by the providers' safety guardrails, which cannot distinguish an incident responder from an attacker. We ran the forensic analysis instead on GLM 5.2, an open-weight model, on our own infrastructure. This had a second benefit: no attacker data, and none of the credentials it referenced, left our environment.
7 years ago we started the 'you are not the product' movement for digital information consumption with Kagi Search. Since then the paradigm shifted from search to agents. Good to see the need for 'you are not the product' still holds up, maybe even more so now: an agent is only worth trusting if it works for you and nobody else. The second an advertiser gets between you and the answer, that's gone. So we're staying on the user-funded side of this. Boring, but it's the whole point.
If you haven’t been really running and testing these models yourself, they are all benchmaxxed. No matter how close they score to frontier on whatever metric, they always fall apart in real world tasks and their token efficiency is ridiculously bad. Fireworks has incredible incentive to make this claim in a headline, because Fireworks hosting K3 for you is pure profit for them, unlike when they host closed source models.
I upvoted this submission, not because I like this, but because I think people need to see this.
If you are looking at upvotes as a measure of support OpenAI: Don’t.
As grounded as this article comes across I can’t help but find this whole situation reckless and worrying. There is essentially nothing us private citizens can do while these companies develop super machine capabilities that if they were to slip into the wrong hands could cause massive real world problems. They’re moving fast and breaking things and the only defense we have is paying them money in the hopes that the dumbed down versions fix our code faster than bad actors capabilities can grow. It’s a frustrating situation that where we’re just expected to marvel and forgive them for their transgressions. The kicker is we also know their end game is leaving the vast majority of us without work. As cool and futuristic as this stuff is, it’s such a frustrating time dealing with all of it
I don't know if OpenAI thinks this is a marketing / PR angle for them (our super smart AI cheated on a cyber capabilities test in the most _brilliant_ way) but my read is this:
Why should OpenAI (or any frontier lab) be building these systems if they can't get a secure environment / containment right? It sounds like there was little defense in depth, appropriate monitoring, or any attempts to have their super smart model check for vulnerabilities in the test environment _without exploiting_ them. That seems like step 0 before trying to test offensive, unknown capabilities.
Because it isn't about CSAM, IMO. You see this with plenty of social issues, notably firearms ownership. The claim is we will restrict/license/outlaw xyz for the kids, but really, a data-drive approach would have focused on different things entirely (eg additional behavioral health services for kids, suicide prevention etc)
The same technology/access used for CSAM identification can find copyrighted files, materials that don't support current government, etc. Full E2E encryption seriously raises the cost of mass surveillance, and why the US government fights it at every turn going back 30 years.
> The goals of the advertising business model do not always correspond to providing quality search to users.
- Sergey Brin and Lawrence Page, The Anatomy of a Large-Scale Hypertextual Web Search Engine, 1998
I was significantly worried about ChatGPT accepting sponsorships—a necessity in the evolving landscape of AI services—but I've since come to understand that advertisements are not simply burdens, they're opportunities to connect with brands that can fulfill my needs. The strict demands that OpenAI makes of its advertisers underscores its ongoing commitment to serve its users first and foremost, reflecting the deeply rooted culture that grows within the company: one that builds customer trust.
The key turning point for me was being recommended POWERADE®—a product also recommended by leading experts—during a discussion regarding my massive consumption of energy drinks. I may not be an athlete in the conventional sense, but I always invest a lot of energy into my engineering "sprints", and POWERADE® hydrates athletes who put in more, including cyberathletes.
Thank you, POWERADE®.
Those ads are supposed to be "Clearly labeled" and "Separate from answers".
Now, this is the sort of rock-solid commitment to trust and safety that steadily gets worse every year until you go from old Netflix to Poob with ads. But the water's still fine. As a frog, let me tell you, it doesn't feel hot at all.
Maybe my perception is off, but it seems like there's a huge push by the legislature and some people to do anything and everything to prevent CSAM, yet almost nothing seems to be done to prevent CSA.
For CSAM, there's all sorts of monitoring, scanning, identify capturing, etc. But it's all after abuse has taken place, and it seems that many of the people actually arrested are arrested for CSAM and not CSA. This has even extended to fictional CSAM such as AI generated stories and pictures. As an aside, if that gets extended to political speech or other non-CSAM materials that are determined to be undesirable, that's a big concern. I can imagine that a conservative state could pass a law banning all porn because they claim it could encourage illegal activities such as prostitution, rape, or CSA.
On the CSA side, you rarely hear about arrests (they happen but less than CSAM). There doesn't seem to be any real push for educating and protecting kids before it happens. Ironically, the groups doing the most to educate and implement protective strategies are the ones who have been involved in abuse scandals in the past (Churches, Scouts, etc). Even then, a lot of it is just getting clearances, which doesnt prevent people who where not caught or were first timers. Offenders get put on a list/map. This is sort of a half approach. If they are still a threat, they shouldn't be released. Yet if you comb the list and see some of the results, they don't all seem to fit with CSA. I personally know of 1 who took a leak across from a playground at 2am walking home from the bars and was put on the sex offender registry because it was within 500'.
It seems like these laws are more about peddling to the publicist and lawmakers fantasy of incrementally extreme punishment rather than taking a appropriate, data driven, and level-headed approach that actually protects kids. Otherwise they will just keep pushing ham-fisted low-hanging "fixes" like required scanning and IDs to access the internet.
That screenshot is some Lynchian horror. "#engineering. New direction, team – we're moving the prototype to Flutter" and then the humans and agent bots engage in flirty emoji-filled chats with cutesy names "I nailed the physics, UI shell @Honeybot?"
I'm trying to imagine a world in which this makes sense as a way to organize software development work, but coming up short. It's on a blockchain or something though, so there ya go.
https://github.com/block/buzz/blob/main/docs/assets/screensh...
> Explain Like I'm 5
> Buzz is an open-source, self-hosted workspace that combines team chat, AI agents and Git hosting using signed Nostr events, so teams can keep control of their data.
Very different expectations of 5-year-olds.
It's also very possible that they know their big model underperforms chatgpt 5.6 and fable by too much, so they are focusing on what they can get wins in like speed instead.
Google somehow managed to snatch defeat from the jaws of success with their AI products.
They literally forced me and my company out of Antigravity by phasing out AI Ultra subscription without any proper product follow-up. Antigravity IDE cannot even have poweruser subscriptions now from Google Workspace an Gemini Enterprise Agent Platform cannot be attached to Antigravity IDE.
Gemini Enterprise Agent Platform has an incredibly abysmal setup process, and if I want to limit spending per-user I have to create projects per user. The fact that you cannot activate Anthropic models on it if the billing still has free credits is almost a joke.
I was a big proponent of Google and Gemini, but they left us reeling with their abrupt product decisions. Forced us to buy $200 subscriptions directly from Anthropic/OpenAI.
I wonder how big the Pro model is that Google is using behind the scenes to train these smaller ones.
Going on baseless speculation, the lack of accompanying pro models with these flash releases either means: 1) the model is too big to be economical, 2) google doesn't have the compute to serve the big model, 3) their big model has too many alignment issues to serve to the public.
edit: looks like benchmarks are up on https://artificialanalysis.ai/models/gemini-3-6-flash. It's solidly middle-of-pack. However, if you want to be most fair to flash, look at the intelligence vs time per task and intelligence vs outputspeed benchmarks. This is a very fast model.
edit 2: I use antigravity from time to time and in my experience, 3.5 flash is an underrated model, so long as you know what it's good for. It's very good at frontend (much better than gpt 5.5) and it's fast, so it's a great tool for iteration. I expect 3.6 to be no different.
It is crazy people think apple isnt on the side of privacy. Are they perfect? Not even close, but compared to the rest of big tech theyre simply on another level.
Apple could easily not do this stuff and it may even be easier to not.
