shon

Originally it was built as the inverse. A signal that recruiters could use to tell them which “passive candidates” could be more willing to change jobs.

A customer asked if it could be used internally (we already had their ATS/HRIS data) so a new feature was born.

Yes money was a motive but this particular feature didn’t seem like an evil idea to be used to increase employee misery.

That said, We did build some things that I do regret now.

We built this tool as part of HiringSolved. Other signals included time in current position relative to industry average and personal history.

If true, probably not for long. Still my point is people are customer. It’s more fun to think about what won’t change. I think we will still have baristas.

This goes way back. Nine Inch Nails was a synth-first band with the music being written by Trent in a studio on a DAW. That worked but what really made the bad was live shows so they found ways even using 2 drummers to translate the synths and machines into human-plated instruments.

Also way before that back in the early 80’a Depeche Mode displayed the recorded drumb-reel onstage so everyone knew what it was, but when the got big enough they also transitioned into an epic live show with guitars and live drum a as well as synth-hooked drums devices they could bag on in addition to keyboards.

We are human. We want humans. Same reason I want a hipster barista to pour my coffee when a machine could do it just as well.

Thanks for the clarification, Ed. That’s quite interesting.

Also congrats on the new company!

GPT4 is lazy because its system prompt forces it to be.

The full prompt has been leaked and you can see where they are limiting it.

Sources:

Pastebin of prompt: https://pastebin.com/vnxJ7kQk

Original source:

https://x.com/dylan522p/status/1755086111397863777?s=46&t=pO...

Alphasignal repost with comments:

https://x.com/alphasignalai/status/1757466498287722783?s=46&...

Agreed. Leadership can sometimes bring actual value ;)

And to be clear, I’m not sure Ed would call himself that. Those are my words, not his.

This great except that the author starts with “our kids love math” and goes on to say how they wanted to figure out a way for their kids to do math everyday because they love it.

My kid loves Roblox. Math not so much. She’ll gladly play Roblox every day with no assistance or structure from me. Math? That’s about as easy to form a habit with as running 5 miles. How do I get her to love math (or other good things) in the first place?

The hard part isn’t getting by kids to do things they love, it’s getting them to do things they DON’T love.

I was able to get her hooked on reading actual paper books but only when she messed up in school and lost access to devices for 9 days. Boredom took over and books were the answer. Now she loves them, thank god.

Interestingly, Ed Newton-Rex, the person hired to build Stable Audio, quit shortly after it was released due to concerns around copyright and the training data being used.

He’s since founded https://www.fairlytrained.org/

Reference: https://x.com/ednewtonrex

Is it? To me it feels like Google is about where Microsoft was in 2002.

Case in point: This thread about using anything other than Chrome…

Wish granted: https://archive.org/details/mosaic-ncsa-evolt_browsers

Perhaps the disturbance you feel is actually the RIAA moving their Death Star into firing range of Stability.ai

What’s your preferred browser?

No, Doom didn’t kill the Amiga. Wing Commander did.

Hmm, thanks for the interesting article. Probably knew that at one time but forgot lol.

Still, it shows my point, that research that reaches too far from the company’s core biz is difficult to recognize as a success. The laser printer, was a better printer and it was an improvement to Xerox existing scan+print business.

Bell Labs also had a lot of commercial success.

PARC, though awesome for society, was a massive commercial failure for Xerox.

As tech nerd, I love the lore of The PARC and went to visit it as soon as I moved to the Bay Area.

However, I assume it would be taught as a lesson of what not to do in a business context:

Investing in pure research often yields innovations that are opposed to existing business lines or simply too far out to see as useful by management.

That said, many companies have research arms. Microsoft, Walmart, IBM, Meta, Google…

I used to despise Microsoft for exactly this behavior. I’ve been more impressed with them lately for their investment in AI and willingness to take risks bringing new products to market.

Unfortunately I’ve had the same problem with Edge after refusing to use any M$ browser for decades, I gave it a try to get early access to “Sydney” which was worth it all until they lobotomized it.

Microsoft at its best, gave us the Xbox by abandoning Windows and focusing on making a good product regardless of ecosystem ties (and using its enormous money/muscle).

Microsoft at it’s worst uses deceptive lock-in tactics on its own users and is still not above force-feeding: “if you want to use this product, you HAVE to use these other crappy products you don’t want.”

This is so strange to me for a company that has worked so hard to transform.

The second one is the larger issue. Patreon is working well for many things but still not easy or integrated enough.

Yep, I stayed there in 2000 to attend HOPE 2000. Traveled with Sir Dystic, Lumi and some other cDc kids. Parties were fun. cDc put on a great show.

Just a reminder that you can relive the good ole days and run your own BBS with doors FidoNET and all sorts of legit oldskool features all in a modern system running in Node

EnigmaBBS:

https://nuskooler.github.io/enigma-bbs/

I love that OpenAI surprised Google and lit a fire under them.

Google’s task now is to think through a post-search experience that includes advertising in a much more useful and intelligent way.

I think it can be done.

This demo makes me think they’re not that far off:

https://x.com/googledeepmind/status/1732447645057061279?s=46...

100% John, I do mean trivial programs.

It’s the stuff that you should not bother an engineer to do, but it can’t be done (in a reasonable timeframe) by a business person or even a technical IT person who isn’t proficient at code.

There’s a space there, a practical need for things that are boring or obnoxious for an actual SWE but that fill some business need or use case. ChatGPT fills these in nicely.

> Making a coffee table out of a solid slab of walnut, for example, will last a lifetime

Not CNC, right?

Yeah I think it's fun to discuss. I think we're saying the same thing. I'm not saying ChatGPT could replace a professional software engineer for most things at this stage. That clearly isn't practical yet.

As you said, it does make mistakes, though it can often correct them if the context window is there. The differences in how we use it for programming tasks and what we find useful is probably related to our SWE skill levels. Yours is likely much higher than mine, as I've never been a professional SWE.

That said, for the many little automation tasks that business people often bug a real engineer for, like iterating through a CSV and hitting an API to do something, then writing a result, it's ability to write simple and useful python scripts has been amazing for me.

I believe it's getting close to the place where I could use to to prototype simple MVPs but it failed during my last attempt.

I disagree with the idea that handwritten software falls into the same category has handmade woodworking. Unlike software, there is no machine that can do the type of woodworking you describe.

At least for now, we still have the unsolved 'robotics problem' where the data to train humanoid robots to do woodworking better than a human doesn't exist in sufficient quantity to make it work using current AI training techniques. This too may soon change via synthetic training data or some modified technique.

Conversely, my own experience shows that even with context window limitations, ChatGPT is already better than 90% of working software engineers for generating small practical programs. Here's why it's better:

1. It's better because it not only knows how to code but is equally knowledgeable about product management, UI design, business practices, etc which makes it more like interacting with an 'architect level' engineer 2. It knows a good if not 'the best' way to do most things

I agree. Reminds me a various sci-fi scenarios: We hit a tech level where jobs and tasks are not done by people unless they want to do them. Or in many cases we still want a human there for some reason. Like the barista, the job could certainly be automated but the hipster human doing the pour-over adds something to the experience that we, as the human customer still want.

I think it turns out that humans will want to work or event need to work in many ways.

I still want to garden (ripping out weeds is oddly gratifying), practice guitar, work on my old Porsche, hack on various projects, etc.

This is a fun sign of the times lol.

Reminds me of the early arguments around automation in racing and whether humans should still be required to shift the car manually once the automatically actuated gearbox became faster and better than a human operator.

In all of these things, humans are transitioning from a place of needing to do work to get the task done, to a place of doing things because they want to do them, because it’s fun, or they learn or it benefits them.

Will people still write code by hand once the machine is clearly superior at it?

lol, that’s terrifying.

Love it OP!

Check out https://github.com/yl4579/StyleTTS2

For your local voice cloning needs.

HN discussion here https://news.ycombinator.com/item?id=38335255

Funny I’ve been talking about cloning my dad for a while in this same fashion. Thanks for the inspiration.

Cool work! I tested it and got some mixed results:

1) it throws an error if it's installed to any drive other than C:\ --I moved it to C: and it works fine.

2) I'm seeing huge latency on an EVGA 3080Ti with 12GB. Also seeing it repeat the parsed input, even though I only spoke once, it appears to process the same input many times with slightly different predictions sometimes. Here's some logs:

Latency to LLM response: 4.59 latency to speaking: 5.31 speaking 4: Hi Jim! user spoke: Hi Jim. user spoke recently, prompting LLM. last word time: 77.81 time: 78.11742429999867 latency to prompting: 0.31

Latency to LLM response: 2.09 latency to speaking: 3.83 speaking 5: So what have you been up to lately? user spoke: So what have you been up to lately? user spoke recently, prompting LLM. last word time: 83.9 time: 84.09415280001122 latency to prompting: 0.19 user spoke: So what have you been up to lately? No, I'm watching. user spoke a while ago, ignoring. last word time: 86.9 time: 88.92142140000942 user spoke: So what have you been up to lately? No, just watching TV. user spoke a while ago, ignoring. last word time: 87.9 time: 90.76665070001036 user spoke: So what have you been up to lately? No, I'm just watching TV. user spoke a while ago, ignoring. last word time: 87.9 time: 94.16581820001011 user spoke: So what have you been up to lately? No, I'm just watching TV. user spoke a while ago, ignoring. last word time: 88.9 time: 97.85854300000938 user spoke: So what have you been up to lately? No, I'm just watching TV. user spoke a while ago, ignoring. last word time: 87.9 time: 101.54986060000374 user spoke: No, I just bought you a TV. user spoke a while ago, ignoring. last word time: 87.8 time: 104.51332219998585 user spoke: No, I'll just watch you TV. user spoke a while ago, ignoring. last word time: 87.41 time: 106.60086529998807 Latency to LLM response: 46.09 latency to speaking: 50.49

Thanks for posting it!

Edit:

3) It's hearing itself and responding to itself...