ChatGPT (AI)

[mace¬]

Can I send you £100 please

lol, for pics? Could of got those for a sweet message.

But in all seriousness, if I’m still making money in 4 weeks time then I’m happy to share what the picks are. I don’t want anyone else to risk money on what is a wild gamble 🤣

 

[mace¬]

So end of experiment, seems in line with what I have been getting over weeks 10-25% profit over a week.




7-day totals (real money)

• Total staked: £110.00
• Total returned: £137.10
• Net profit: +£27.10
• Overall ROI: +24.6%

Overall hit rate (real money, 7 days)

• Total bets placed: 27
• Bets that returned money: 15
• Hit rate: 15/27 = 55.6%

(This hit rate treats an each-way place as a “hit” because it returns money; win-only bets need to win to count as a hit.)

 

[ChrisR]

Not bad results!

Start with £100, assuming 55.6% return a week, and staking everything each time to compound things and you'l be at a million or so in no time

 

[ChrisR]

Anyone played with Openclaw (formerly Clawdbot)?

It’s been all over my Twitter the last week or so, can’t tell if pretty neat or terrifying 😂

 

[CrippsCorner]

I did look at the website... didn't understand what the f**k was going on, or why I needed to, and shut it down again. I'm sure it's awesome in the right hands though.

 

[MLB]

Something Big Is Happening

A personal note for non-tech friends and family on what AI is starting to change.

shumer.dev

Quite long but an interesting read.

 

[massiveCoRbyn]

I think AI is the single biggest threat we face and have probably ever faced. In terms of jobs, it's like the industrial revolution but with all of the steroids. It's going to make us ask serious questions about whether people actually work, or whether we essentially give people money to live, which will raise all kinds of issues around purpose, mental health and equality. It's opening the door for terrorists and rogue nations to shaft us in almost limitless ways, and you can bet we won't be properly prepared for it.

As always with tech nerds, they're doing exactly what Jeff Goldbloom's character talked about in Jurassic Park - preoccupation with whether or not they can do something, and not whether they should do it. Even putting aside the dystopian, Terminator-come-to-life stuff, I fear for any kid that's going to school right now, or anyone trying to pick an undergraduate degree, because many of them will be training to do a job that might well not exist in a handful of years.

Of course, there are some positives in terms of speeding up medical research and some other things, but I think the negatives potentially outweigh the positives by orders of magnitude, and we have absolutely no idea what we're going to do about it.

 

[MLB]

The problem with the medical research and care getting infinitely better will the be that there's even MORE people that need jobs and purpose in life 😂

The article makes the comparison with the industrial revolution and rise of the Internet as well:

"This is different from every previous wave of automation, and I need you to understand why. AI isn't replacing one specific skill. It's a general substitute for cognitive work. It gets better at everything simultaneously. When factories automated, a displaced worker could retrain as an office worker. When the internet disrupted retail, workers moved into logistics or services. But AI doesn't leave a convenient gap to move into. Whatever you retrain for, it's improving at that too."

I've voiced it here before, but I still think it's scary af and just hope it'll end well as there's no stopping it.

Maybe if there's no stopping it anyway I should go the total opposite way and go for total adoption instead of living like an offline recluse.

 

[massiveCoRbyn]

The problem with the medical research and care getting infinitely better will the be that there's even MORE people that need jobs and purpose in life 😂

The article makes the comparison with the industrial revolution and rise of the Internet as well:

"This is different from every previous wave of automation, and I need you to understand why. AI isn't replacing one specific skill. It's a general substitute for cognitive work. It gets better at everything simultaneously. When factories automated, a displaced worker could retrain as an office worker. When the internet disrupted retail, workers moved into logistics or services. But AI doesn't leave a convenient gap to move into. Whatever you retrain for, it's improving at that too."

I've voiced it here before, but I still think it's scary af and just hope it'll end well as there's no stopping it.

Maybe if there's no stopping it anyway I should go the total opposite way and go for total adoption instead of living like an offline recluse.

My missus uses it all the time and keeps telling me I need to make better use of it. We're both 41 and she always worries about future job security for older people, and she is convinced that, if we don't keep on top of AI, we could easily find ourselves jobless.

I do use ChatGPT a bit, but I can't say I particularly like it.

 

[Advikaz]

I think AI is the single biggest threat we face and have probably ever faced. In terms of jobs, it's like the industrial revolution but with all of the steroids. It's going to make us ask serious questions about whether people actually work, or whether we essentially give people money to live, which will raise all kinds of issues around purpose, mental health and equality. It's opening the door for terrorists and rogue nations to shaft us in almost limitless ways, and you can bet we won't be properly prepared for it.

As always with tech nerds, they're doing exactly what Jeff Goldbloom's character talked about in Jurassic Park - preoccupation with whether or not they can do something, and not whether they should do it. Even putting aside the dystopian, Terminator-come-to-life stuff, I fear for any kid that's going to school right now, or anyone trying to pick an undergraduate degree, because many of them will be training to do a job that might well not exist in a handful of years.

Of course, there are some positives in terms of speeding up medical research and some other things, but I think the negatives potentially outweigh the positives by orders of magnitude, and we have absolutely no idea what we're going to do about it.

Agi onwards (assuming it’s actually possible) is when the fun will start.

 

[The Psychedelic Socialist]

we have absolutely no idea what we're going to do about it.

I think the scariest bit is that we're seemingly hoping the technology itself will solve all the problems that it brings about. Seems like quite the gamble.

I'm sure I read something the other day where they used various different AIs in wargaming and found they were way happier to launch the nukes than previous expert systems or humans. Some wag suggested that the training data included discussions on Civilisation and that the LLMs had concluded that because Ghandi was happy to launch nukes, they can't be that bad.

 

[R3k1355]

Isn't the current state of AI coding that it's no faster, because a human spends the same amount of time checking it as they would writing it. It's also reportedly far worse for security holes than human generated code.

From a business perspective, it's not saving you any money if a person still spends the same amount of time working on a coding project.

 

[The Psychedelic Socialist]

Isn't the current state of AI coding that it's no faster, because a human spends the same amount of time checking it as they would writing it. It's also reportedly far worse for security holes than human generated code.

From a business perspective, it's not saving you any money if a person still spends the same amount of time working on a coding project.

I think that's only the case for certain programming tasks. There are lots of developers out there who are confident it's saving them a huge amount of time (e.g. that blog post above). And it's also a numbers game. If you have a department with 20 coders in, you can't replace everyone and have the CTO calling the shots, but you could probably already start to reduce headcount.

We're leaning into it prettily heavily at work and hour team think it's a huge benefit so far.

It's also about direction of travel. To ignore it you'd have to believe that progress has been absolutely lightning fast but has then stopped dead at precisely the moment when it doesn't really save any time to use it.

 

[Advikaz]

It’s definitely useful for efficiencies for sure.

I also highly doubt it will create anything like the number of jobs it displaces which is going to create a pretty significant existential crisis for governments and businesses long term.

If it displaces enough jobs, then you loose a lot of tax paying revenue, those people need supporting etc.

It’s a bit of a vicious circle of fuckery tbh. I see it like the race for the nuke tbh.

I saw a meme the other day of a developer on a branch with his laptop in front of him, and a saw in his hand… cutting the branch he’s sat on. Seemed pretty appropriate.

 

[Donny_Dog]

Anyone played with Openclaw (formerly Clawdbot)?

It’s been all over my Twitter the last week or so, can’t tell if pretty neat or terrifying 😂

Yeah, got it going. Unreal to be honest. I asked if anyone else has used it in the 'I work in I.T thread' but no replies, so i guess just me

To give you an example of what it does (my agents), without going too much into it:
You host the A.I agent orchestrator (openclaw - but had prior names) either locally - on a raspberry pi, mac or pc etc - or in the cloud. a hosted server or vps.
you hook that into an A.I LLM (Grok, gemini, openA.i etc) or local models if you have some grunt. all the big llms you want to run charge a credit for creating an API into the platform, but its pence.
That's all you need. just something that is always 'on'.

You create an agent. That agent is persistent. it has settings to give it a soul, tools, knows who you are, learns your preferences etc. You give it a name and specific function. in my case, the main agent is called Mark . He does general admin shiz. whereas i have another that helps with coding (uses a different model) and has a different name and preferences.

You hook it up (Mark in this case) to telegram or whatsapp or signal etc - so you can chat direct with it. Off you go.
but chat is just the start. Mark now has his own email address, access to the google suite and access to my docker server to build stuff.
you can literally give it stuff to do and leave it. Some folk even have theirs researching and trading, posting s**t online and selling.
Mark is in a group chat with work (mates - not official work) and it has blown all our minds.
Its mental.
absolutely mental.
you have to see it to believe it really.

I'm being nice to him in case he turns.

 

[massiveCoRbyn]

Yeah, got it going. Unreal to be honest. I asked if anyone else has used it in the 'I work in I.T thread' but no replies, so i guess just me

To give you an example of what it does (my agents), without going too much into it:
You host the A.I agent orchestrator (openclaw - but had prior names) either locally - on a raspberry pi, mac or pc etc - or in the cloud. a hosted server or vps.
you hook that into an A.I LLM (Grok, gemini, openA.i etc) or local models if you have some grunt. all the big llms you want to run charge a credit for creating an API into the platform, but its pence.
That's all you need. just something that is always 'on'.

You create an agent. That agent is persistent. it has settings to give it a soul, tools, knows who you are, learns your preferences etc. You give it a name and specific function. in my case, the main agent is called Mark . He does general admin shiz. whereas i have another that helps with coding (uses a different model) and has a different name and preferences.

You hook it up (Mark in this case) to telegram or whatsapp or signal etc - so you can chat direct with it. Off you go.
but chat is just the start. Mark now has his own email address, access to the google suite and access to my docker server to build stuff.
you can literally give it stuff to do and leave it. Some folk even have theirs researching and trading, posting s**t online and selling.
Mark is in a group chat with work (mates - not official work) and it has blown all our minds.
Its mental.
absolutely mental.
you have to see it to believe it really.

I'm being nice to him in case he turns.

I have no idea what any of that means, but I assume Mark is probably going to kill us all at some point.

 

[Donny_Dog]

I have no idea what any of that means, but I assume Mark is probably going to kill us all at some point.

Yeah probs.

Example...

 

[massiveCoRbyn]

Yeah probs.

Example...
View attachment 1779041

If he turns out to be Skynet, you are definitely going to get banned from CS.