Earlier quoted context omitted.
> Whatever you can do to create deterministic gates that the LLM simply must overcome to reach a “done” state, do it. First thing Gemini did when I tried that was turn off all the rules in eslint.config.mjs claiming they were "overly stylistic" Yes, it got better once I explicitly told it not to disable any rules, so I accept I was holding it wrong but I do worry just how many footguns it puts into other things becau…
Why does your harness allow the LLM to generate output that can disable rules & checks?
Zuckerberg says AI agent development going slower than expected
591–600 of 661 posts
Re: Zuckerberg says AI agent development going slower than expected
#592Second thought, I wonder if zuckburger has turned into the Cramer of tech. First, the metaverse fiasco, now another massive overinvestment into a dead-ish end?
Re: Zuckerberg says AI agent development going slower than expected
#593AI should have caused a job market boon: because less skilled employees would have been more hirable/useful. That this is not the case leads me to suspect that AI is an excuse to reduce employee count, but not the root cause.
Re: Zuckerberg says AI agent development going slower than expected
#594Earlier quoted context omitted.
How much of your money do you spend on paying for kids school lunches, paying medical bills of terminally ill kids and paying off the student loans of graduates? It's very easy to say that someone/some oeganization's wealth should be confiscated, yet I have yet to see those proposing it actually putting any of their own money where their mouth is.
> How much of your money do you spend on paying for kids school lunches, paying medical bills of terminally ill kids and paying off the student loans of graduates? A third of my salary each month.
Re: Zuckerberg says AI agent development going slower than expected
#595If AI was a productivity boon: wouldn't a company employ the same or more employees to capture more market share at a competitive edge. Shedding employees because they are more efficient seems like shooting yourself in the foot to try and stand in the same spot in a race. AI should have caused a job market boon: because less skilled employees would have been more hirable/useful. That this is not the case leads me to…
Re: Zuckerberg says AI agent development going slower than expected
#596Earlier quoted context omitted.
Do I write all my code with an agent now? Yes. Can you just give an agent a desired outcome and let it work, unsupervised? Absolutely not. I strongly suspect that developers moving from writing code to managing agents to write code for them is very similar to developers moving into leadership and management roles and managing ICs to write code for them. Some devs just 'get it' and thrive, leading a team really well a…
I disagree with your 2nd assertion. Even engineers who are less tech lead style engineers, can gain a significant boost in productivity by being able to quickly run through POCs and build an understanding of surrounding areas of their work, so they are able to contribute more. For eg I am able to make React changes much faster and the changes are higher quality, given frontend dev has never been my job role. I’m able…
Man, if I had a dollar for every time someone said "I'm not good at X, but LLMs are so impressive at it". Like do you think there might be some connection between those two points?!
Re: Zuckerberg says AI agent development going slower than expected
#597Earlier quoted context omitted.
It's a concrete measure of available resources. More so than revenue.
It is precisely the opposite of a concrete measure of anything but investor valuation. Investor valuation (or even just asset value, for that matter) doesn't have much weight when figuring out how seriously to take a CEO's words. Revenue at least has a modicum of signal of competency.
And this is Meta we are talking about not Tesla, it's P/E is like 20 something. The valuation is very reasonably correlated to the earnings.
Re: Zuckerberg says AI agent development going slower than expected
#598If AI was a productivity boon: wouldn't a company employ the same or more employees to capture more market share at a competitive edge. Shedding employees because they are more efficient seems like shooting yourself in the foot to try and stand in the same spot in a race. AI should have caused a job market boon: because less skilled employees would have been more hirable/useful. That this is not the case leads me to…
This technology isn't even a decade old. It hasn't even been really useful until the last 3 years or so. Why would you expect it to be transformative already?
"Internet is just a fancy fax machine" stuff.
Re: Zuckerberg says AI agent development going slower than expected
#599Earlier quoted context omitted.
Do I write all my code with an agent now? Yes. Can you just give an agent a desired outcome and let it work, unsupervised? Absolutely not. I strongly suspect that developers moving from writing code to managing agents to write code for them is very similar to developers moving into leadership and management roles and managing ICs to write code for them. Some devs just 'get it' and thrive, leading a team really well a…
> Some devs just 'get it' and thrive, leading a team really well and building a great culture. I don't think it's this because the outcome you get from AI isn't controllable. You can give it the best prompts and design suggestions and it'll still give you completely wrong or horribly written code. If you were a manager and one of your reports kept producing completely wrong and horribly written code that other folks…
I don’t have a dog in this fight but it seems you’re not accounting for iteration and feedback. A horse will veer off of a road if not occasionally nudged to stay on it, but is useful transportation, nevertheless.
Re: Zuckerberg says AI agent development going slower than expected
#600Earlier quoted context omitted.
for "the rest of us" building certain kinds of systems on top of it. as integrated into Meta's platforms it is clearly uncompetitive as a consumer chatbot with the likes of ChatGPT, DeepSeek, Microsoft Copilot, and Gemini. Like the other chatbots give me useful answers to questions even if they're wrong sometimes. Meta AI, on the other hand, cracks jokes about my foxwork. That's no so bad because sketch comedy is one…
Ok, I'll bite. What's "foxwork"?
https://www.amazon.com/dp/0231133383
and found out that people really do it. It took me about 300 days to contract with a community of foxes and started "going out as a fox". At first I didn't want to explain it to people at all but when I botched my explanation and my wife gave me some tough love about "cultural appropriation" I realized I needed a cover story and it is "I go out as a character to do street photography and make people smile"
And it became real. My son didn't believe it until he saw me work: my son saw these two college girls walking down the street and thought "there is no way they would talk to me" and next thing he knows they flagged me down and are asking me questions and I took their picture. I went to fireworks in Groton, NY and the next day in Little York [1] I haven't even parked my car and people I saw the day before are excited to see me.
When I get approached every day it becomes effortless and automatic to make an approach and when I feel like I am getting a 50% take rate I think "I am doing something wrong and I need to regroup", my usual take rate is in the 80-90% range but it feels like 100%. I don't even talk with other street photographers about it on forums because I'm drawing from an entirely different probability distribution.
Thanks to this work, it looks like I will be teaching about magic soon [2] so I was working on a short monologue about blessings on the drive up to Little York which primed me to really respond when people I photographed said simple things like "Enjoy the Fourth of July" and eating soft serve ice cream from a vendor felt like the feast that Martin Prechtel says is the only ritual -- and I am still feeling high from it just like I felt high from going to Litha put on my my pagan friend.
See https://mastodon.social/@UP8/tagged/foxwork
[1] one of the five Yorks of New York
[2] Magic is real, full stop. Science is real too, and there is nothing more scientific than the double-blind test that accounts for the placebo effect which is as I see it: "the patient experiences the treatment as receiving a blessing and feels better" and "the doctor experiences the treatment as giving a blessing and perceives the patient differently"