HNHacker News
TopNewBestAskShowJobs

ygjb

2,392 karma · joined September 8, 2011

Yvan Boily Founding Security Engineer Co-founder, Oghmere parentsguidetoai.ca https://nimblemyth.ca
submissionscomments
ygjb··on Canada will match US tariffs 'dollar for dollar' as trade talks break down
That might have been the case two years ago, but there is no shortage of decent open source solutions that meet or exceed most business use cases. I am sure some enterprising consulting firm will come along and offer integration of those + Canada located inference infra + OSS models (Chinese, or otherwise). I know that there is at least one Canadian inference provider, Augure, but I haven't tried them yet, as well as a few actual model developers in Canada.
ygjb··on OpenRouter is joining Stripe
The first half yes, the second half, I don't think so, because that would require OpenRouter to know what you are charging your customers.
ygjb··on OpenRouter is joining Stripe
A single policy location for multiple models with usage limits/rate limits, spend limits all configurable using a common syntax + deep integration with account levels and high visibility into subscription/payments related to meter usage to help understand whether or not your pricing model is underwater as costs change is exceptionally valuable. AFAIK each product consuming model providers without an openrouter style proxy has to do that work themselves, and if you are already using stripe for subs/payments, the direct integration is a huge value add for customers, especially if you can alert on "upcoming price changes from your provider changes your unit economics for XYZ packages/sub tiers".
ygjb··on Claude Code May–August 2026 weekly limits promotion
One of my challenges is that I often end up burning tokens trying to build the right solution rather than building a solution that is good enough and deferring the right choices until later. I have been actively working to change my expectations for building with AI to accept worse solutions to get things going rather than trying to solve everything at launch.

This has been a recurring problem for me - as a security engineer catasrophization is a fundamental skill to finding vulnerabilities in complex systems, but makes me to conservative when building. For some of my leaders they are much better at saying 'good enough, ship it, and fix it later'.

ygjb··on Claude Code May–August 2026 weekly limits promotion
I'm ok with that. When the.major AI players implode under debt, it will create a massive proliferation of folks who go on to start new businesses unencumbered by the debt and bad decisions by the current leaders.

The collapse of the massive overvaluation and circular investments will be the undoing of several major tech investors and companies that have a serious contribution to the things in tech we don't like. It will be bad for the economy, but it will also weaken the abhorrent control that those companies have over regulation in the United States.

It's gonna be a rough time, but it's pretty clearly necessary.

The only big problem is that if that collapse happens under the current administration in the US, they will be utterly incompetent to respond to a real domestic crisis, since they have been virtually unable to do anything without creating more problems.

ygjb··on We Tracked a Shipment of Rare Books. It Ended at an Amazon AI Training Facility
I would rather make the case that with the increasing consolidation of media under the big tech companies, regulation is more effective than a business case.
ygjb··on Google wins bankruptcy auction for Spirit Airlines emails, chats, documents
I would love to see a rider from the courts mandating a significant third party test by privacy researchers to see if the deidentified data can be reconstructed from models post training.
ygjb··on Ask HN: Alternatives to GitHub
Contributions aren't just commits/prs. It's also opening issues. The harder it is to provide feedback, the more likely users are going to seek a different library or tool, and depending on the actual objective of releasing or publishing code, that is not ideal.

I do agree on the fly by night agents, especially since so much is actually just agents acting on behalf of folks who just spam whatever permissions are requested in Claude/Codex.

ygjb··on We Tracked a Shipment of Rare Books. It Ended at an Amazon AI Training Facility
First, to be clear, I work for Amazon, and I don't work on this or related efforts, I am a security engineer. I also won't talk about or answer questions at work, and my comments are more generally about the practice (all of the major organizations building AI are doing destructive book scanning). These are my opinions, and do not reflect my employers (past or present).

You might be right that they can't do that now. The simple path forward is to have these companies simply make a public commitment to publish the data when the copyright expires. I also think there is a space to be carved out, probably through regulation, to ensure that there is a clear path for these scans to enter the public domain, at the very least.

This is not just important for these specific books, I have written on other platforms and in other spaces about the importance of media companies and those who benefit from strong copyright laws to protect and generate profits and revenues to repay the public for the cost of that enforcement over time by ensuring that at the appropriate time, those works fully enter the public domain. That could mean a restructuring of the Library of Congress in the United States to become a modern Library of Alexandria to host data, and shifting to a registered copyright model where to gain the protections of the court, you need to upload/submit your copyrighted works for storage and eventual release. I doubt it would ever happen there because of the amount of money invested in tying up IP in the United States, but perhaps a more amenable location like the EU could help with that.

However it might work, part of the promise of the Internet was that information would be liberated, but we see every day how much information gets sent down the memory hole when businesses, sites or services shut down, or how regularly companies abuse IP related regulations to attempt to strangle competition. It would be expensive now, but it would create an incredibly valuable legacy of information for the future, and it will only get more expensive to build such a thing as time goes on.

ygjb··on We Tracked a Shipment of Rare Books. It Ended at an Amazon AI Training Facility
It would be really great if the companies who are doing this would commit to placing the scanned files into a public trust that would coordinate with organizations like the Gutenberg project to ensure that the scanned materials enter the public domain on schedule. Publishing encrypted archives with the keys in escrow would be a good first step.

IMO that would go a long way to resolve any concerns about losing books. I still don't like the idea of extremely hard to find or last prints being actually destroyed for this, but it certainly makes it more palatable.

ygjb··on Anthropic becomes the 'Apple of AI': Most revenue despite being most expensive
UI, polish, and customer experience are innovation. It takes an enormous amount of work to build consistent user experiences and build products that make choices on behalf of users that don't alienate those same users, yet Apple is able to consistently do that in ways that produce fervent customers who love their products.

I personally don't like or use most Apple products outside of my work MacBook pro, but I recognize the value and innovations there.

ygjb··on Is the industry ready for tokens-constrained work?
I routinely hit the limits on a Claude pro max 20 account for personal projects. For context I have 26 years of professional experience and have been programming for around 40 years.

In addition I have virtually unconstrained usage at work (for now).

Tasks that I do on my personal account: - building stuff for my wife's business - analyzing and experimentation and building harnesses to test other models/services - processing interesting research papers that publish without code or sufficient data to replicate the work

And that's besides simple stuff like exploring new topics I am interested in, and building learning assistants and personal tooling, and eliminating technical chores.

ygjb··on Ask HN: How do you go from writing code to deploying with agents?
I use a combination of automated and manual testing. I spent a bunch of time setting up playwright testing at the start of my project, and then I shunt that through ZAP as well so that it can have a good understanding of how things work.

If gives me a pretty solid map of what is working well, and aids in driving through security testing.

Still not a replacement for manual testing. Us meatbags still have a knack for royally messing up the expected use cases :P

ygjb··on Silicon Valley misreads science fiction and undermines democracy
https://www.youtube.com/watch?v=WOSqCjMRXWA Granted, the wealthy may not have the resources of the US military, but I am willing to bet that push comes to shove, the corporate security forces of companies like palantir and anduril, let alone actual mainstream military hardware providers would be a far greater threat than armed civilians in a fully deregulated or collapsed American state.

I don't care much about the 2nd amendment absolutists, and the myth of the American heroes who will stand up to tyranny, because so far the only thing they appear to have stood up for in the last hundred years are gun rights, religious fundamentalism, and in 2020, a demagogue.

American exceptionalism and gun culture are important to Americans, but are pretty much viewed as narcissism and with skepticism outside of the US. YMMV, but like another comment I made recently, this isn't an anti-American sentiment, it's just a perspective from one of your neighbours to the north who has lots of Canadian and American friends and family living in the US.

ygjb··on Silicon Valley misreads science fiction and undermines democracy
I said resources not money. Charisma, religion and promoting hate have empowered just as many psychopaths in history as wealth.
ygjb··on Silicon Valley misreads science fiction and undermines democracy
Jokes on you. When there isn't a functional system of government there is nothing to stop rich psychopaths from paying poor psychopaths killing those who get in there way and declaring themselves dictators.

In the absence of effective laws to align people in the direction of the society they want, and constant effort to realign in response to pressure, we will fall back to tribalism and all of the attendant forms of feudalism and government that came before it. Ownership then reverts to might (or those with the resources to buy might) makes right.

ygjb··on Qwen3.8 Max now ranked as the best overall model by agentic index
Because that doesn't satisfy the narrative of American exceptionalism. It's easier to point at something and say it was stolen or copied than it is to compete, especially with the political climate in the US.

This isn't an anti-American sentiment. It is an anti-corporate/regulatory capture/embrace and extinguish sentiment (which probably reads the same to many people these days).

ygjb··on DeepSeek planning to significantly raise prices
Ah yeah, but that's not an AI problem in general. The models are trained on human language and human centered expression. I realized after re-reading, that my comment was more harsh than I intended, I really was trying to be helpful. The first step of asking for help should be asking if you have framed the question or task correctly, that way you build your own understanding of what you are asking the model to help you with. For better or worse, under the hood, the computer is a machine following instructions that largely appear non-deterministic, so if you don't frame the inference request correctly, it just boils down to a GIGO problem.

It's probably not what you want to hear, but are you sure you were holding the tool correctly? At least with this tool, you can actually ask it :D

ygjb··on DeepSeek planning to significantly raise prices
It's not really possible to help without more information about the models you are using, the harness you are using, and how you use the harness.

You haven't even provided enough information for us to know if you are using the right tool. Saying "the model" in relation to an LLM provider that offers several is like asking for help with using a Dewalt or Milwaukee to help assemble an cabinet. Are you cutting, drilling, hammering, screwing?

ygjb··on Google fixed more Chrome bugs in June than over the past two years, thanks to AI
Sure. Wealth has always given early adopters and first movers wings.

As hardware specs get better, and model providers keep making different classes of mistakes that alienate users, more people will opt for or support local models. I tend to take Schneiers' "Attacks only improve" philosophy and apply it to damn near everything like this. Inference will get more expensive, but IMO cloud based inference is at or near the maximum that users can tolerate, across multiple spectra. There may be a marginal increase in cost, but as the different metrics between what is affordable for rapidly scalable, cloud based inference and the lag of local inference with open models and expensive hardware converge, prices will come down.

I am not the right person to say what the peak is going to look at, but when the companies with (practically speaking) unlimited compute resources and money are starting to take a good hard look, there is going to be a drive towards efficiency. I hear grumbling at work, and the appreciation from my leaders when I show up with a tool that shows how I reduced my token costs, or I keep asking the questions of what the token cost (real, and internal billing) are for tools my peers write are.

I also have absolutely eye-watering personal AI bills that are starting to make buying a better tier of local inference gear look more affordable, even with inflated hardware prices. The current generation of open models are very effective, and paying for multiple $200 dollar a month subs is not feasible longer term, and as one example, in one month, if I was paying enterprise token rates, I would have burned through $8000+ token budgets on one of those accounts, doing actual, practical useful work, it makes more sense to buy some much better hardware than to pay for one-time inference costs.

ygjb··on Google fixed more Chrome bugs in June than over the past two years, thanks to AI
Until a couple of years ago you needed to be able to read research papers and code to build fuzzing harnesses and triage the findings to find vulns. Now you can point Claude, codex, Kimi, or even some smaller open models that run local and get a servable fuzzing loop in a couple of hours.

That's democratization - shifting power and skills away from those with the wealth to purchase vulns and run offensive security teams. It's still not simple, but it's one of the reasons bigger vendors want to lock down the capabilities of models that present a threat to the wealthy and powerful.

ygjb··on Google fixed more Chrome bugs in June than over the past two years, thanks to AI
Yes, sort of. Not that there is anything wrong with that. Fuzzing yields tons of trash results, the occasional fleck of gold and the even more rare nugget. Then you analyze how it found the flakes and nuggets update your fuzzing harness. Every so often you get to write a paper or give a talk about a breakthrough to share meaningful improvements.

Current AI isn't super effective at making the breakthroughs, but it sure is effective at democratizing the ones you can point it at.

ygjb··on Microsoft Can Track Users via a Windows Device ID
That puts the OS in the position of attempting to profile or determine if an application is accessing OS, hardware, and user details to build a fingerprint, vs using those capabilities to do something the user intends, which puts the OS developer into a performance sucking, soul sucking arms race against big and little brother surveillance/advertising platforms. I absolutely support the intention, just know that it's a brutal battle :(
ygjb··on Fable 5 Is Back
Yeah Claude does this for me all the time. I have a template project I use that also leverages puppeteer/webdriver/Firefox, and I can point Claude at the template and a website and it will happily build me an MCP service that it can use to interact with the site if there isn't an API or MCP already available.
ygjb··on Fable 5 is Back
Are you sure it refused because it can't use a username and password? I literally have loops running right now where it uses a database of test users and passwords to log into different roles and do computer use and browser automation testing. Sonnet and Opus complain when I provide credentials and password in chats but it is happy to use ones stored in files and stuff, so it might just be guardrails to push good opsec so that the secrets aren't captured in the session history and prompts.
ygjb··on GLM 5.2 beats Claude in our benchmarks
It's pretty simple. There are things that I do because it's fun, like gamedev. I hand code that, and don't use LLM tools because I like learning and building. I do lots of utility stuff coding for my wife's business, most of that is stuff I could do in a few hours. It's worth $20 to not spend a few hours doing it. It's a cost benefit tradeoff. I won't learn much fixing WordPress themes or adding a feature to her web page, or setting up an automation for her, so I don't see the point of doing that.

Same thing for stuff at work. Oh, the tables/schema changed and my queries broke? I could dork around with spark and cypher for an hour, or I can tell claude to update the queries for the new schema. At the rate I am paid, spending on Claude tokens is generally a better use of my resources.

Building a net new solution? Coding tools take a back seat until I get the core logic right, then I let automation handle web page and UI scaffolding.

ygjb··on European Commission lines up Amazon and Microsoft for cloud gatekeeper status
Why?

The US likes to flex it's muscle by pushing other governments around on behalf of its commercial interests.

I think it's well past time for Canada to either formally pursue toes with the EU or pursue alignment with EU consumer protection legislation.

Middle powers unite!

ygjb··on VibeThinker: 3B param model that beats Opus 4.5 on reasoning with novel SFT+GRPO
> Could you teach a 5 year old to drive a car? A 10 year old? A 12 year old? To drive a car requires being able to read, to have judgement about ice or rainy conditions, to anticipate a child running after a ball. By the time a human in in their mid teens they have acquired the base knowledge...

I would be interested to see a formal study of this. I say this not out of anything other than a observation that I think the only real blockers are a) judgement, and b) physical reflexes/strength. As a kid I was certainly aware of ice,snow, and rain, because I road my bike year round and had low confidence in my own ability to control my bike on snowy or wet terrain, especially during season changes. That translated into learning to drive in northern Canada in the winter and applying those lessons to driving.

In an environment devoid of consequences, I have seen kids operate driving simulations (both real simulations, and video games) with a degree of precision that is shocking, including seeing several 9-11 year olds play the simulations and games with a much higher degree of confidence than adult drivers. Children have an awareness that the simulations are consequence free, unless given other motivation. Adults that are consistent drivers have muscle memory and preconceived expectations that govern the decisions they make when playing the game. I am curious about the level of training and exposure required for children to overcome their lack of awareness of the hard limits and consequences of driving and driver error, versus the amount of training and exposure required for expert drivers that are novice gamers to stop applying their learned experience to consequence free simulations.

ygjb··on Ask HN: What are you working on? (June 2026)
https://parentsguidetoai.ca

I spend all day working on AI related stuff at my day job, but more recently I have been getting lots of questions about AI from friends and family and that ramped up when the Vancouver school board announced they were rolling out Copilot across high schools.

I put together the guide, made a few iterations and published it and now I am getting more questions than ever and it's turned into a blog too.

I'm having fun but it definitely feels like a topic that is underserved. A solid example, lots of resources for when your kid is the victim of a deepfake, but what if you as a parent know your kid is getting into trouble and escalating. Where do you go for help to make sure they don't escalate from unfriendly behavior to abusive or even criminal.

It's a big challenge!

ygjb··on Statement on US government directive to suspend access to Fable 5 and Mythos 5
Shh...you'll burst the bubble of the folks who think that LLMs are toy stochastic parrots...
← PreviousPage 3 of 24Next →