Hopefully sooner rather than later we get some kind of compute function on Autonomi.
1 Like
When we do, will @19eddyjohn75 demand 1 nano == 1 Teraflop? 
Sorry Eddy, I don’t mean to crush your dreams.
5 Likes
I don’t really feel like my dreams are getting crushed, whatever economic model is chosen, is ok. I know what my 1 nano is worth, it’s way more worth than whatever people have on their bank account.
I sympathize, with the bankers their Stockholm syndrome… 
5 Likes
I want to expand the 1 nano concept to everything. 1 house, -car, -food, -tech and everything else, 1 nano.
Never have something so small been worth so much, 1 nano FTW. 
0,0000000001 
The paradigm shifting new economic model I suggest it will be named ONE, ONEism and so on.
The new economic model will also be conspiracy friendly so it will be a good fit for Tyler and similar.
2 Likes
And a practioner will be known as an onanist.
Suits you, sir.
1 Like
My phone just vibrated to tell me that
https://minepi.com/
are looking for partnerships.
DYOR but maybe worth an hour or so of somebody’s time.
Allegedly 60 million mobile users.
1 Like
3 Likes
Don’t listen to me, what do I know, but then again…
2 Likes
There will probably be two groups of people, those who jump on the AI/LLM’s train towards the future and those who stand on the platform watching as they get left behind, age is probably a huge factor in the decision deciding if jumping onboard the train or not.
1 Like
The UK is done but still the crazed “left wing” will demand more immigration, more taxes, more support, and decry anyone who opposes this as evil and “an extremist”.
The irony is lost on them, but I am not fleeing.
1 Like
- use their experience to shape expectations and validate those expectations using evidence from others to determine whether/when it becomes worth adopting
- those who go ahead because it feels exciting and worthwhile regardless of 1.
Option 2 Is fine, I do that still as an old man - just not in this case because my assessment is that there’s a lot of snake oil being sold here. So 2. should always be tempered at some point by 1.
The snake oil appears to be off-the-scale so I prefer to dabble, watch and review in accordance with 1. Age has little to do with this. I’m as excited and interested in technology as I was when designing and building stuff using the what then were cutting edge electronic components - microprocessors, ASICs and TTL. There was no need to question this, positive evidence was all around. Since I’ve seen a lot of snake oil take people on wasted journeys and quite a lot of that has been in AI going back to the nineteen eighties.
4 Likes
Worst news I have heard today.
Go on, go on, go on. Flee!
3 Likes
I think it’s a question of what this tool is able to do and how it can be used in a reasonable manner…
… Efficiency gains (as I’ve heard numbers of +70% for programmers) are imho highly questionable…but it’s nice to just ask for the correct syntax of something and get the answer in the same format always (+in my mother tongue…) …
But e.g. Definining the right regex for my use case and remembering how to give names to different patterns in search strings are found pretty fast with chatgpt and have cost me many hours in the past…
In our company it seems we come to a place where we don’t send every word to a professional translation service but ai is pretty good at modifying our json files to integrate new languages…
… So I think there is use cases where it does make a lot of sense… And there is this stupidity where my company seals a multi million dollar deal because we write in the contract that we do something ‘with AI’ … Without anything specified about performance/acceptable error levels/anything relevant…
Ps: best use case in my life is writing mails from rough drafts - and for that the local ai is way better than chatgpt because I can include private info and don’t risk forgetting to replace placeholders…applying for childcare places/a new rented home…
3 Likes
I agree, I have found one reliable use for it so far but it is trivial and arguably no quicker than a web search, and if I’m doing it regularly not as good as learning the syntax I’ve forgotten or not yet learned.
A problem is how it is being sold. Another is how often it is used by people who think or feel it is a good idea, but don’t understand the reality and unintended consequences. Another is that it helps “level up” those who don’t know quite how to write good code, which is a very wrong assumption at this point but feeds on wishful thinking by them and those employing them.
There are more, but it seems likely that relying on programmers thinking that Copilot is helping is not a good measure, and that’s just one instance. I anticipate many issues with the tech, even for people how already know what they are doing.
2 Likes
Some new AI models are recursing now - and to toot my own horn, I predicted they would implement such a long time ago in another thread here.
Recursing allows a sort of pseudo-thinking and by implementing methodological staging to this process, so in the first level work to build more structure into the query, breaking it down into steps, then processing that, and then recursing over it again to look for errors and hallucinations, it comes out with much better output.
I think the ChatGPT o1 preview is doing this and I posted about another one in this thread a couple of weeks ago.
All of this is becoming more representative of human thought. The cost is that it takes a lot more processing power per query. Hopefully more limited and specific models + ASIC processing (I posted about a company - Etched - in this thread a week or two ago as well) will allow this cost to be pushed down dramatically.
1 Like
2 Likes
https://awala.network/ popped up on Hacker News today. I only had a quick browse, but I wonder if it’s in any way useful or interesting to the Autonomi team or to any networking-lovers here.
1 Like
Far more radical and actually working advances in computing, and this is from 2018. This is something that could be changing the world by now, but for some reason isn’t being given billions.
That might actually be a good thing in the long run because the last thing we want is for real world object communal computing technology to be owned by mega corps.
This is 43mins and possibly the most innovative thing in computing since the Microprocessor unleashed computing on the masses and created the software industry.
This could literally bring computing literacy to the masses in the way the teaching everyone to read and write brought a world of knowledge and new ways to discuss and create to everyone.
42 minute presentation by Bret Victor (ex Apple UI/UX specialist):
2 Likes