j-conn
j-conn on Hacker News
Nice. I found myself doing a lot of repetitive stuff handing off between agents, time to fix that. I think I agree with your approach of keeping it lean.
I also sometimes rewind the conversation after going off on a tangent, sometimes not if I want the agent to have the sorts of things I’m thinking about in context. I probably justify keeping too much crud in context than I should
What do the session hooks do?
“The bug is a memory corruption issue in WeChat's VoIP stack. We're withholding the technical details for now. We plan to present the full analysis at an upcoming conference.
This specific WeChat bug is one instance of the many unconventional attack surfaces that are present across many messaging apps. We're conducting more of this research across other apps and attack surfaces, while working with app developers on attack surface reduction. This may take an industry-wide effort, since some of it depends on the platform owners. Once that work is further along, we'll share our progress, including the technical details of this WeChat bug.”
So deepfakes and “AI was also used to generate synthetic identities that bypassed KYC checks, craft hyper-personalized phishing lures tailored to individual victims, and produce AI-driven malware capable of evading signature-based detection.”
Our most vulnerable must be getting absolutely hammered by scams. Is anyone working on protecting e.g. elderly people from social engineering? I feel like banks should be proactively suggesting people have a second family member or fiduciary sign off on risky transactions like wiring money to new accounts or pulling out a ton of cash.
I was recently on the phone with a very articulate member of my banks fraud department, or at least I believed they were up until they asked for my two factor code to reset my account. They had visibility into some past legitimate wire transfers which they asked me about along with a fictitious fraudulent one that they “blocked“, and when I pushed back against giving them my two factor code, they had a whole script prepared about needing to freeze my account until I went to a branch with four forms of identification (or just give us the two factor code and avoid the hassle). An acquaintance wired their down payment to a scammer after the real estate firms email system was compromised and scammer inserted their own wire instructions.
Not to mention reports of mythos directly hacking into banking systems. Scary times.
My only point was that this article shouldn’t be considered authoritative, wanted to put it in perspective for someone surfing hn and just reading the comments
The person closest to me was saved by CPR after cardiac arrest (and cooling at the hospital), with no neurological deficits
Totally misleading. Early CPR (+AED if available) absolutely saves lives. Article is from 2011 by a family med doctor.
Overly aggressive resuscitation attempts are definitely a problem but context matters
Citation for the phone, WiFi, Bluetooth claim? That’s an extraordinary claim and I’m very skeptical.
The “fusion” mode looks really cool. Also loved the logo grid under “for the paranoid” — looks like a customer list but except it’s a wall of shame (“they did not run bromure”). Thanks for sharing!
What are some of the implications? Where does widely available mythos-level hacking lead? By people with a vested interest, do you mean non-cloud software vendors?
Incredible! I’m interested in doing something similar on windows, have you looked into that at all? Apparently codex computer use plans to support this on windows in the future. Were you able to see how codex was doing it, or the inspiration was just “they’ve shown it’s possible”?
> ‘AI’ doesn’t exist, and LLMs have vanishingly narrow legitimate justifiable use cases. … I’m yet to identify any application of a LLM which can rationally be mistaken for intelligence.
What’s the opposite of AI psychosis? Burying your head in the sand? Because anyone who could write this unironically today is certainly afflicted.
Which do you like? Barebells salty peanut and chocolate dough over here. Though the sugar alcohols certainly aren’t great for you either, I think they were recently linked to stroke risk
Have you tried docker sandboxes?
https://docs.docker.com/ai/sandboxes/The docker desktop license requirement is a factor, though. You need a paid subscription if your company has something like 250 employees or $10 million in annual revenue
(EUC = electric unicycle)
All the more important for people who value the rule of law to continue to call it the Department of Defense. Executive power ultimately depends on acceptance and compliance. Corrupt and unconstitutional actions are only legitimized by collective acceptance
100% agree, many people don’t realize just how harmful wood smoke is. It’s also the main source of pollution in the Bay Area during the winter. Unfortunately energy costs are high enough here that people resort to burning wood to save money, so collectively beneficial policies are likely to face resistance (understandably).
The purpleair map has been awesome to at least make the problem visible. I hope they are using it to aid enforcement on spare the air days.
In what world is he a “pseudo” philanthropist? He’s already given more than $60 Billion away and pledged to donate 99%+ of his wealth. He has also called for his class of ultra-wealthy to be taxed more
Not everyone is so unimaginative and myopic ¯\_(ツ)_/¯
There was. From the article:
'''
ZeroEyes said that trained employees review alerts before they are sent and that its software can make a lifesaving difference in averting mass shootings by alerting law enforcement to weapons on campus within seconds. At Lawton Chiles, the student flagged by ZeroEyes was holding his musical instrument like a rifle, co-founder Sam Alaimo told The Washington Post.
“We don’t think we made an error, nor does the school,” Alaimo said. “That was better to dispatch [police] than not dispatch.”
'''
“It’s not rocket science — you’ve either got to pay more, or you’ve got to let in way more people. … There are wonderful, caring people all over the world who would like to come care for our seniors at the wages we’re willing to pay, and we just have to let them in,” Gruber said.
This is the crux of it. The government should also subsidize and directly administer more senior care, especially given the economic drag from having family members step into these roles
I was really surprised this is technically possible on iOS, so I watched the demo video again and noticed the hardware plugged in and a “screen broadcast” going on. The dongle is mentioned [here](
https://www.ycombinator.com/companies/blue):
‘On iPhone, a tiny USB-C plug-in (“Bud”) provides precise input control’
Really cool they got this to work and it seems great for accessibility. I wonder how Apple feels about it
Mind sharing your detailed planning workflow and/or how you came up with it? Is it basically a detailed prompt asking it to produce a PRD?
I often find my plans for the code changing as I get into implementation. I wonder if giving the model explicit permission to change the plan would work or cause it to go off the rails for big chunks of work like this.
Any specific products you’d recommend for this?
I’m not sure about that mechanism, but the impact on blood sugar is quantified as “glycemic index” and “glycemic load.”
One thing to keep in mind in the US is that dietary supplements aren’t regulated by the FDA. Consumers have to place a lot of trust in the manufacturer that their product is free from potentially harmful contaminants.
Very cool, thanks for elaborating! Looks like that alpha has an improved (github flavored) markdown implementation too
Widespread testing makes contact tracing viable, assuming a big suppression in cases from lockdown.
Mind sharing what features in particular make it hacker friendly? By page scraping, do you mean how it fetches the page title when you paste a link, or have I been missing out on some awesome advanced feature?