mywittyname

> And now your friends and coworkers will send you AI generated mail anyway

This hits close. I realized one of my friends was using AI to message me and I took it kind of hard. It's weird to be worth the effort for them to set up a chat bot to talk to me but not worth the 2-3mins a week to actually read/respond to my messages.

Right now, I just basically ghosted him, but I have teh feeling this is the start of an emerging issue.

I think they understand, but they are also building machines that need to run Windows + pre-installed bloatware without being so obviously bogged down no the sales floor that no one buys them.

Por qué no los dos?

This is a flawed argument, the current system gets rid of good performers all the time, and we have evidence of wage suppression, so employees don't get to negotiate on fair terms. You just get stuck in your pay band.

I agree it would be a good counter-factual, but I think the differences would be more around industry stability. Particularly, I think the ability for employees to push back against historical threats like off-shoring would have made the industry more appealing to younger people looking for something stable, and prevented this weird cycle of labor shortages causing salaries to explode, unqualified candidates pivoting to the industry using low cost training solutions (bootcamps, shitty masters programs), then companies failing to deliver on initiatives because the people they hired are poorly trained.

If we had 30 years of steady growth in CS education, then we'd have more experts in the field, doing a better job at executing. And it would likely cost companies less in wages as well. There are many industries where incredibly talented people make fairly modest salaries while producing world-changing products.

This is the fruition of Microsoft's dream, since it's the most obvious way to drive copilot usage in a way that A) burns mad tokens, B) is actually useful to paying customers.

Though, I have to wonder if distracting leadership with shit like this will be bad for business in the long-term. Both because leadership will fail to do their jobs, being too busy playing peeping tom on employees, but also because it takes their eyes off the prize - measuring the things that make money.

Go tooling has this kind of thing as well. I'm not a huge fan of go, but last time I had to work with it, we leveraged a lot of codegen.

There have been plenty of projects that do send junk data to these endpoints. The problem is the junk data gets users noticed because some manager looking at dashboards gets an alert about some supersonic Corolla driving down the Pacific. And they go yell at the team responsible.

As a result, analytics endpoints generally have some authentication and verification built into them. Obviously, with enough time it's possible to reverse engineer these components. But that's a lot of time and effort vs just blocking the request.

Just be aware, this is something that will be noticed. I've been building analytics systems for a while now and have had people do this. Usually it gets picked up by the anomaly detection system or as an alert in the ETL pipeline when we try to transform it.

Personally, I just plop it into a "dead letter office" table, then verify it's not malicious. But it's possible other companies would handle this differently.

Out of curiosity, I checked if gcc would optimize i = i++ out, and it does!

Question for you, since you seem like you know what you're doing here: How do I learn about stuff like agents.md? I feel like every week there's a new Claude feature people are talking about and I have no idea how they learned about it.

Yeah, the solution was to link it to the nvidia page of the card, then it was like, 'oh, okay.' But at that point, I lost faith in it's ability to provide me with the information I was looking for. If it's information is so out of date that it doesn't know about the 5000 series, how could I be confident that it knew the details I was asking about (game engine related research)?

> As much as I hate to admit it, step one in most of my projects now is to ask AI about it. Maybe it’ll tell me something I don’t know.

Or, more likely, it will tell you something it doesn't know.

Reminds me of yesterday, when I was arguing with ChatGPT that the 5070TI was an actual video card. It kept trying to correct me by saying I must have meant a 4070ti, since no such 5070ti card exists.

Also, doesn't musk hate him? I have to imagine he's behind this.

Why are you miffed about it? I legitimately hate reading golang with passion and find python to be pretty intuitive, outside of the odd ambitious list comprehensions. I worked in a golang shop for several years, so it's not just an familiarity situation either.

We are just different. That's not something to be mad about.

My parents hate AI, as do my nieces and nephews.

It's not an age thing. It seems to be an experience thing. If AI is perceived to have ruined something you like, or otherwise negatively impacted your life, you're going to hate it.

> I'm not sure about that. The "normies" around me love AI

Really? Because every "normie" I know complains to me about how much they hate AI.

Lots of the the complaints center around it being forced on users, how all information on the web is AI generated anymore, thus can't be trusted, broader issues around how it makes society much less stable (bots, misinformation, fakeporn, and more recently, breaking security). And, of course, those whose jobs are actively being replaced by AI really hate it.

The data centers going up locally are likely to be another avenue for complaints.

I disagree. Nothing about this was slow at all. They wrote endlessly about how they would throw the frog into a vat of lava, dangled the frog over the lava for a good while before tossing them in.

FB Marketplace is full of scammers too. Every time we've tried to sell something, we've had people try to steal our phone number and register it with their Google Voice account.

They do, it's only like 1-2 million barrels a day in capacity right now.

> How do you see AI creating more jobs? (I need some hope right now lol).

More armies of one. That single team of five now becomes 30 teams of one or two each.

Which largely how automation resulted in more jobs - the cost decrease induced demand. Think about how cheap cameras, laptops & internet up-ended traditional media. We went from 3-4 channels on TV in the 60s, to 3-400 channels on cable by the 90s, to 115 million channels on YouTube right now. Because anyone with a basic phone can record and edit content, which used to require millions of dollars in equipment and took years to learn to do. And people are happy to do so for a fraction of the revenue a TV station would require.

Because it's the law misunderstanding technology.

Chatbots are not people. They are computer programs. And there's no other realm I can think of where merely interfacing with a computer program breaks attorney-client privilege.

It is equivalent to saying an email to your lawyer breaks privilege because you communicated with gmail. And it gets turbofucked when you consider that a program may be sending your information to an LLM. Would this same judge rule that having copilot installed in Outlook also breaks privilege because they "chatted with an outside party" while drafting an email (even if they didn't intend to send it to copilot)?

I can't think of a reason this isn't about the technology.

Well, calling Claude a "third-party communique" here is the stretch.

Say a person used Excel via Office 365 to run some calculations to be given to their lawyer for their defense. Is that considered to be "communicating with a third party?" I don't think so, it's just a computer tool.

We call them "chatbots" and anthropomorphize LLMs, but, despite the name of Claude's parent company, Claude is not a person.

You can use an offline model via ollama. I'm sure better tools will emerge for less technically-inclined individuals.

Seems like there might be demand for chat clients with end-to-end encryption.

It's significantly different. It's better quality, more accurate, and more convincing.

That probably means they receive a lot of small donations. Payment processors often have a fee structure that's 2.9% + <flat fee around $0.30>. So any donation below ~$4.50 would end up having a >=10% processing fee.

There could be currency exchange rates that are factored in at the donation end as well.

I agree that 10% is high, but it's still explainable.

I kind of wish countries would just define, "terms of service" for everyone and not allow companies to modify them further.

I think the takeaway is consumers want to trust the products they are buying are as advertised.

Even just having the agent write scripts to disk and run those works wonders. It keeps the agent from having to rebuild a script for the same tasks, etc.

It's been demonstrated that some companies, even F10 ones, have been using pirated content to train their AI.

I'm not sure why people hate using the query string so much.