Notices by PointlessOne :loading: (pointlessone@status.pointless.one)

@wwahammy are any of the devs even in California?

I remember illegal numbers (it was not that long ago actually) when people went out of their way to meme in the face of law.

@thomasfuchs Depending on how you define “outsourcing and offshoring” it kinda sorta worked. It still does. You can totally find an agency-type of company to build you a mobile app of a website and the quality is not even bad, sometimes even good.

We collectively dank on Indians but among them there are a plenty of competent and even brilliant developers. Likewise among local devs there are a lot people producing substandard output, for any definition of “local”.

I myself arguably “an Indian”, not actually Indian but I live in and work from a country most people wouldn’t find on a map until 4 years ago (and probably to this day). I'm the guy on the other side of email/slack/zoom. Throughout my career most of my colleagues seen me in person for only a couple of weeks total, a good fraction—never. I’m proud that most of my work was valuable for my clients and that most of the time our partnership was a net positive (mostly by a large margin) for them.

There absolutely were botched “outsourcing and offshoring” but the scale of the damage and rectification potential are incomparable to what's going of with LLMs.

I don’t think @searls responds to or even reads replies on here but…

I find his piece on Ruby Central a bit odd.

https://mastodon.social/@searls/115282174363254250

It has some new to me information but I don’t see how it’s relevant to the whole thing.

For one, it’s all about André. Reads a lot like an attack on his character: see how not very good he was a decade ago, he must be not good at all now.

The issue is that the RC incident involves a lot more people: HSBT, Marty, RC board of directors, the whole maintainers team both present and historic, and—indirectly—the whole Ruby community.

Even if everything Justin written turns out to be true, and even worse than he states, it doesn’t address anything that actually happened this month in any shape or form.

We—conveniently—don’t get a similar piece on anyone else directly involved in the incident.

@fasterthanlime It's some sort of host monitoring API. It can do port checks, HTTP checks, pings.

@nikitonsky I think this is actually a good design. With a transparent thumb you get to use the full width of the widget for the range of values. You can clearly see what the value is. You also can get a bigger thumb. This allows for easier interaction as the action target is bigger.

With a solid thumb you have to either lose precision (it’s hard to see where the thumb center is when it’s a big circle), range space (range didn’t give the whole of widget to accommodate thumb), or reduce action target size (if you make thumb smaller to fixe the previous issues).

Of all the issues with liquid glass this here is not one of them. I think this is actually an improvement.

@nikitonsky happiness is the side effect of memberberies overdose.

The whole second half is all memeberberries. Painstakingly recreated envs and imagery of the first movie with barely 1% of new stuff. It doesn’t instill confidence that we’ll get much of anything original here. And if we don’t why not rewatch the original instead?

@catsalad OMG! I urgently need a clear png of this.

@raine @JessTheUnstill

@rysiek @lyncazu yeah, I know. I just don’t have an x acc and the heat map itself most likely comes from a paper anyway.

@rysiek @lyncazu maybe this then?

https://arxiv.org/pdf/2311.14737

@lyncazu @rysiek maybe this?

https://arxiv.org/html/2405.17399v2

@eniko Except they still ship LLM integrations in all their OSes this year.

@inthehands I don’t think it’s entirely fair to put full responsibility on the “reply guy”. OP knows what kind of platform they post on. They are aware that there will be a Reply button under their post.

If OP doesn’t want a reply they should use a medium that doesn’t provide an opportunity for reply. A blog with disabled comments, a static web page, or even an offline medium like a paper journal. Or be ready to ignore all replies to that post.

We can blame platform’s affordances all we want but we can not deny that the choice of platform is entirely on the OP and they are at least partially responsible for the outcomes of that choice.

@inthehands @twsh I’m not convinced this framing is useful.

I agree with points 1 and 2, and even accept 3 as a logical conclusion. But there’s a counterexample: AIs produce coherent language. It’s often hard to tell the nature of the author of a piece of text produced by a LLM, sometimes impossible. So from the reader’s point of view the text says things. The fact that AI has no communicative intent, or that it has no coherent world model is irrelevant.

This is not unique to AI. People on a regular basis gleam meaning that was not intended by a speakers/writers.

@viq @ryanc @rysiek AI even without quotes is a marketing term. We have a perfectly serviceable term for useful AI systems: Machine Learning. ML reflects the nature of the thing much better, in my opinion. We throw a bunch of data at a machine so it learns the patterns and produces similar output. There’s very little intelligence in matrix multiplication.

@ryanc @rysiek Well… Waking up The Great Dreamer to be consumed first is basically the basis of AI accelerationism movement (they call it themselves).

And to the point about safety: it’s basically the main argument of AI doomers (they don't call it themselves but seem to get into the name as of late). It's important to get AI right the first try because we might not get a second chance. Where AI is often AGI (which is a stated goal of OpenAI and Meta, at least) but doesn't have to be. Bringing an existential threat onto everyone might be not as easy as 7 words but all these AI systems are released as if it's completely impossible.

@ryanc @rysiek If that's your wish I hope it’ll come to be but what if… it decides to spare only you and worship you as you're technically its creator? These days it's not easy to get what you want from an AI, not on your first prompt, anyway.

@ryanc @rysiek And that’s, kids, how Skynet became conscious.

@h3mmy @skaverat I guess the question refers to how some systems treat missing data as 0 and if this has any effect.

There was a few stories about similar things.

A guy got a license plate NULL and started getting tickets to random people from all over his state.

Another person got harassed for stealing phones because tracking system put a pin on map at their house. Turns out the system didn’t have more precise data than a country so it put the pin at a geographic center of the USA which happens to coincide with that house.

Also, Null Island. https://en.m.wikipedia.org/wiki/Null_Island

@mattly
I am not a supplier: https://softwaremaxims.com/blog/Not-A-Supplier

Google operated on the social contract of providing traffic to sites in exchange for content. Now that #Google is switching to #AI answers synthesized from the content and effectively cutting traffic is there even a point in participation?

Google will keep bringing load to the sites without providing any traffic.

I suspect we are about to see first robots.txt snippets banning Google crawlers like other AI bots really soon. And eventually outright banning Google ips altogether.

I wonder whether we’re also about to see a surge of old-school search engines. With modern technology it’s feasible to run a small search engine from a living room (e. g. https://kukei.eu) on an off-the-shelf hardware and open source software. It doesn’t take the whole Google to operate a decade-old (feature wise) Google search engine. It might be feasible to run a good search engine on small resources that can be paid by simple ads or small subscriptions.

https://www.platformer.news/google-io-ai-search-sundar-pichai/