didibus

18 ms·

361.
by didibus 1y ago
I think Public Benefit Corporation (PBC) are a way to try and solve that problem. I recently switched to Kagi and their Orion browser, and that's when I learned about PBCs. A PBC legally takes a triple mandate, the first is just as any
362.
by didibus 1y ago
So many times you write a document for someone to review and question that it could be written better. Yet the reader will now just ask for bullet points from the AI. I was hoping people would go, right, let's just start writing bullet
363.
by didibus 1y ago
Does your company have a stake in AI? I suspect a lot of companies that go that route are pushing a marketing effort since they themselves have a stake in AI. But I'd love to hear from truly customer only businesses, where AI is pure c
364.
by didibus 1y ago
I'd say the only issue what the "too dumb, but whatever" comment. That should have been removed and it would have been totally reasonable. Edit: Oh actually, the author has another blog post titled "I'm too dumb for
365.
by didibus 1y ago
I've seen Zig popup a lot recently. What's the value proposition of Zig? It's not immediately obvious to me. Is it kind of like the Kotlin of C, going for a better syntax/modern features but otherwise being very similar?
366.
by didibus 1y ago
It's not brute force search, more like stochastic sampling of random variants. It's a directed search where feedback from the environment prunes and weights the search space, and reproduction is a stochastic process biased by fitn
367.
by didibus 1y ago
I love KDE, I wish it was the more common and popular Linux desktop over Gnome. It's really usable and efficient, works great. If it wasn't for Mac laptops insane battery life, performance, quality of build and trackpad, and amazi
368.
by didibus 1y ago
> Leetcode mostly relies on memorizing patterns Math is like that as well though. It's about learning all the prior axioms, laws, knowing allowed simplifications, and so on.
369.
by didibus 1y ago
If you work at "product manager" level, that would be vibe coding, as you prompt for functional and non-functional changes and you review the behavior and characteristics of the program, not the code generated. I believe the autho
370.
by didibus 1y ago
Agentic already implies what you want to imply. Maybe just call it agentic coding? I feel that be the most clear. Agentic coding implies any workflow using AI agents. Which mean it's always the same agentic coding loop: 1. Prompt
371.
by didibus 1y ago
I want to understand the distinction you're making against vibe coding. In vibe coding, the developer specifies only functional requirements (what the software must do) and non-functional requirements (the qualities it must have, like
372.
by didibus 1y ago
Isn't the biggest bottleneck just that they need to adequately and reliably be able to do useful work at a better price-performance ratio than a human ? And that's just not the case yet?
373.
by didibus 1y ago
I think I understood it as the database will basically store data in a binary format that can be fed into the GPU directly, and will also be optimized for streaming/batching large chunks of data at ounce. So it's "optimized f
374.
by didibus 1y ago
It sounded only OLAP from the article.
375.
by didibus 1y ago
Can someone be clear about what this is? Just MCP support to their CLI coding agent? Or is it MCP support to their online chatbot?
376.
by didibus 1y ago
That's a good point, and I'd assume it comes down to what people consider the least of two evils, possible corporate espionage or maybe blackmail, versus CSAM (assuming you believe monitoring for CSAM will help reduce it). I was m
377.
by didibus 1y ago
I admit to not being a museum head myself. Now that I'm a parent though, I've gone to them all, multiple times. Before that, I'd not gone to any of them unless they're world famous. If it wasn't for kids, nobody wou
378.
by didibus 1y ago
I think the issue is that the EU believes to do these things above corruption. In a sense, if you think you are upholding human rights, free speech, and personal privacy, you don't think it is required to offer people ways to hide from
379.
by didibus 1y ago
I remember reading that there was an active re-framing of the computer worker. It was somewhat reflected by employers that started hiring exclusively men and of computer workers being associated with nerds and geeks, and the combination cou
380.
by didibus 1y ago
> I'm not sure why this would be any different. The issue here is you need to make a trade. It's not like cutting out alcohol. Now you have to decide, what alternative painkiller will replace it. There was an initial reason why
381.
by didibus 1y ago
Ya, an increased risk of 5-20% on an already very low risk. That means mothers who don't take Tylenol have baseline 3% chance their child will be diagnosed with autism. And mothers who took Tylenol (at the levels of the study) may have
382.
by didibus 1y ago
Off course, 100% prediction accuracy cannot be achieved. I just mean that, if you're an ML scientist team, you don't just go, we got 76% accuracy, let's close shop, mail in your resignation, job over. From that angle, it&#x27
383.
by didibus 1y ago
I agree the model is predicting language and not actually running the math. That is a point I try to stress too. It is not thinking through a problem, it is predicting what text would look like if someone were working it out. But the traini
384.
by didibus 1y ago
When tuning predictive models you always have to balance precision and recall because 100% accuracy is never going to happen. In LLMs that balance shows up as how often the model hallucinates versus how often it says it doesn’t know. If you
385.
by didibus 1y ago
I agree with everything you said except: > Trying to eliminate cases where a stochastic model the size of an LLM gives “undesirable” or “untrue” responses seems rather odd. Take it back to what it is like you say, this is a predictive mo
386.
by didibus 1y ago
The word "hallucination" mis-characterizes it. LLMs predict the likely tokens to follow the context. And they can make incorrect predictions. LLMs therefore don't have perfect accuracy of prediction. When their predictions ar
387.
by didibus 1y ago
Can't agree more. > If you give them a task and spell it out, they can knock out code for it at a really good pace and wow upper management. This is so true. I sometimes spend entire days, weeks, all I do is provide those type of en
388.
by didibus 1y ago
> and they're allergic to automation, hiring higher quality engineers, and hiring more engineers to improve velocities I think there's another issue, but it could relate to your first two statement here. Even to try ideas, to e
389.
by didibus 1y ago
> so the anti-vibe-coding people feel like they need to claim that development speed has no impact at all Strange, I'd been more of the impression that this is an argument from pro vibe-coders. As more data comes in, the "produ
390.
by didibus 1y ago
Oh wow, the argument was not expected but also very compelling, I'd been wanting to quit commenting as it's a big waste of time and often leaves me sour, as you almost always comment in disagreement or engage in tough debates and

More ›