bazmattaz

Welll yeh until a Hargrove fails on your self hosted server

Yes 100%. I see this all the time. You’re asking about a product and the LLM will cite a source from a competitor where the competitor will review the source and list a few positives about the product but lots of negatives. Then the LLM uses them in the response. So cheeky

Is the problem that the response sounds like AI? Or the response is incorrect? Or too verbose? Or a mix of all three or something else?

What if the response was perfect and sounded exactly like your colleague. Would you care if it was AI generated then?

I think there could be a world where a colleague could just ask its (finely tuned) AI tool of choice to respond for it and colleagues would hardly notice as the output would be so good!

I can’t help but think that that world would be so much more productive as we would spend less time in slack answering messages and more time doing more important work.

However there are two very important things that need to be in place for this to happen: - colleagues have to have their AI tool know all of the their context on their projects….etc and and in-depth understanding of how they write - colleagues must review (and tweak if necessary) the output

My guess is that the problems you are seeing is that your colleagues are failing on one or both of these points

I guess compatibility hell is why Linux will never go mainstream. Users like their favourite apps just working on their favourite OS.

It’s not a great experience if only some apps on supported on your favourite Linux distro while others aren’t.

Could you not be extremely clear and state which distros are supported and make it clear that it may work on others but that they are untested

I personally don’t think there is anything wrong with this. To the critiques I would say; this is the world we live in now. There are LLMs capable of essentially perfect writing skills. We need to get used to seeing a lot more content either written by or finished by LLMs.

The best practise for writing docs with LLMs in my opinion, which you have done, is to write as much as you can first then feed that into an LLM for context, and then work with the LLM to finalise it. Maybe half the time is spent writing and half the time is spent going back and forward polishing the doc.

Finally I think it’s important to give the LLM very clear writing guidelines based on your own writing style. I did this by feeding Claude around 20 of my handwritten docs and asked it to analyse my writing style and then add thy to its Claude.md. After a free round of iterations you can get great results!

I’m just curious, why do you think it should be open source. It’s a private company. The source code is their IP

Same for me. I’m not an engineer (but worked with them for 2 decades) and AI has been amazing for me self hosting.

For example I could never setup Traefik correctly because I just found it too complicated. Now I have Claude I finally got it setup just the way I want it - the ROI on my Claude subscription has been off the scale!

The obvious downside is that I might not really know what exactly I’m implementing and why. I do read all the explanations that Claude gives but it’s hard to retain this information. So there are pros and cons to relying on AI for this kind of stuff I suppose

I find ChatGPT so infuriating the way it always agrees with everything you say. The product is optimised for engagement so it wants its users to be delighted

Dam this is so accurate. As a project manager turned product manager this is so true. You need to estimate a project based on the “pedigree” of your engineers

I’m not an engineer and now I realise why I’ve been struggling getting OpenClaw setup in docker. I just can’t get it to work. Makes sense that it needs access to the underlying OS

Thanks you. Not sure what an acronym was needed for that

Anyone else feel that it’s exhausting keeping up with the pace of new model releases. I swear every other week there’s a new release!

I have the same stakeholder requirements so we use Trello to track everything in our house

This is really cool. At the moment we use a Trello board to track all of our house jobs, projects, reminders and issues. Then we have Home Assistant for all of our sensors and such.

My only concern about something like this is the WAF. My wife uses Trello. She would likely not use terminal. Also she only uses a phone so lack of mobile app or website is another blocker.

But it’s a cool app

I found it quite depressing to read. This guy spent so much time to put just one offender behind bars but the are likely hundreds of thousands out there. So sad

My manager mentioned that his manager (an executive) is not happy because the org we are in are not using as much tokens as other orgs in the company. Pretty wild

My guess is that this is going to be the future for LLMs too. It will get harder or more expensive for AI companies to train their models on the latest information as most sites will block the scrapers or ask for a fee.

There might be a future where you’ll have to pay more for an up to date model vs a legacy (out of date) model

Huh? Explain

100% agree. I think the subscription should be more like $30/$40 per month based on the value you get but of course users would never pay that!

So I wonder if there would ever be a middle ground between what artists would accept and what users would pay?

I think it’s a bit disingenuous to say that Spotify “cut” the revenue per stream. Spotify doesn’t have a fixed revenue per stream, payouts are dynamic and based on how much revenue it made and how many streams there are in total. Also Spotify gives something like 80% of its revenue to the music industry.

It’s essentially a pie that gets divided up amongst all artists by steam count. If your friend’s garage band are angry that they only get 0.001 cents per stream then they should be angry at Spotify listeners for streaming too much Taylor Swift!!!

Furthermore to pay the artists Spotify has to pay the labels and of course they take a giant cut. This trope of Spotify being an evil corporation is unfair

Spotify is a profitable business now though right? Sure it doesn’t make huge profits due to paying out about 70% to larks but it’s finally profitable

Why do you say so? They’re burning cash and don’t currently have a viable path to profitability. Sure they have millions of MAU but not many paying ones.

From what i understand, and I’m no expert, is that all the smart money Is able to get in on an IPO early on way before retail investors have a chance. By the time the retail investors have a chance it’s too late and usually the stock dips post IPO as the early investors sell.

Of course the stock can rise again eventually.

‘Empire of AI’ maybe?

The to be honest for me toggling on and off the VPN every time I want to watch YT is just too annoying. I’m happy just watching YT as free in safari with ad block installed.

I read that if/when Google find out that you’re not actually Albanian they’ll immediately band your entire account. So I wouldn’t recommend doing this trick for your main Google account that has your email and Google drive stuff….etc. you likely want to create a burner account just for this.

Also do you connect to the Albanian VPN all the time? That would mean all the websites you connect to now think your in Albania

Gmail is the worst for this

My favourite part is when it opens a form or something to fill out but if you navigate away from this in-app web view to another app and back (say to get a password) you lose the session. It’s incredibly frustrating

I’ve been using Ad Block Pro with safari on iOS for awhile and works perfectly. Even blocks YouTube ads (I watch YT via Safari)