677: I Accept the Battery Costhttps://atp.fm/677
-
@USBTypeSTeve Estimating AI water usage is not straightforward https://www.youtube.com/watch?v=H_c6MWk7PQc But if/when we do enter a "water crisis" (really the "world's first"?) I'm currently inclined to blame climate change (and politics) rather than AI.
@siracusa @USBTypeSTeve We could start by eating less burgers: https://bryantresearch.co.uk/insight-items/comparing-water-footprint-ai/
-
@vmachiel Like electricity, AI will (eventually) live or die based on the ratio of its usefulness to its harm. Crypto mining lives on because it’s useful for crime and blackmail and speculation, but it seems like it will never reach the mass market in the way that, say, smartphones have.
I think AI is already more useful than crypto, and it also has a much better use-to-harm ratio. Your opinion depends on your own estimate of those use and harm values.
@siracusa @vmachiel I think the issue with this line of thinking is that the usefulness of AI comes mostly to rich software dudes in the global North (that’s all of us in this conversation) while the harms disproportionately go to those in the global South.
As long as the harms don’t come for us, we won’t care, and when they do, it will be too late.
-
@jamesnvc The only “compliance” required will be your continued use of technology products and services. Even today, there are few tech products and services that are not using ethically dubious LLMs without your knowledge or consent. That’s why I say “barely a choice” today.
@siracusa @jamesnvc you still have a choice in how much you use them though. Just because there are restaurants serving the flesh of murdered animals everywhere does not excuse anyone from “having a choice”.
I think you’re trying to move the conversation from “should we be doing this” to “let’s examine what happens as a result of us doing this”, which I totally respect as a way to move the discussion forward, but I don’t think the framing of inevitability is accurate.
(This is coming from someone who eats animal products and uses LLMs — I’m not trying to establish some moral high ground, just trying to remain somewhat objective).
-
@siracusa @vmachiel Energy usage is also often overstated. Nobody is arguing that AI does not consume energy, or that aggregate demand does not add up. But this has to be evaluated in relation to the use and that it is spread across hundreds of millions of users. When broken down per individual and compared with what a typical Western European or American uses energy for, AI usage is negligible.
@secundus
I would be interested in references if you have them. I’ve been having a hell of a time finding numbers on this stuff. -
@fds @atpfm I’m always here for the fahrenheit rant
I don’t think anyone is working on taking it away from @caseyliss - we can have both
-
@caseyliss @fds I’m Danish which is still European last I checked (Trump hasn’t invaded us yet
) and I will be happy to defend your right to use Fahrenheit as much as you like as long as you let me use celsius since I prefer/am used to that 
-
677: I Accept the Battery Cost
https://atp.fm/677If you really don’t like AI, we have some bad news for you.
@atpfm Using a web browser from the biggest advertising company in the world seems like a bad idea. Safari 4EVA.
-
@marcoshuerta I agree with all of this! Uber is a similar recent example: "Lose money until we've destroyed enough of the existing market that we can start charging higher prices and customers have nowhere left to turn."
The relative timing of the bubble-pop (or even more mild "consolidation") vs. the rate at which these tools improve and we learn how to use them well will probably make a big difference in how this goes.
@siracusa Claude Code like products are one of the most token-intensive applications since massive chunks of code (and essays) are being sent into the context window over and over. That makes profitability even harder - Anthropic started charging Cursor more and Cursor had to raise prices. (and both still lose money!)
-
677: I Accept the Battery Cost
https://atp.fm/677If you really don’t like AI, we have some bad news for you.
-
677: I Accept the Battery Cost
https://atp.fm/677If you really don’t like AI, we have some bad news for you.
@atpfm "We don't have a choice. This is here. This is happening. It's happening with or without us…You can be a part of it or you can fall behind…You kinda have to get on board."
@marcoarment making an excellent defense of Tim Cook getting on the Trump train
-
@atpfm "We don't have a choice. This is here. This is happening. It's happening with or without us…You can be a part of it or you can fall behind…You kinda have to get on board."
@marcoarment making an excellent defense of Tim Cook getting on the Trump train
@intarga This is the same defence everyone's always said, and was rebutted by John mentioning that everyone can make their own decision that aligns with their morals and values.
The argument has always been that people do not agree with Tim Cook's morals.
-
677: I Accept the Battery Cost
https://atp.fm/677If you really don’t like AI, we have some bad news for you.
@atpfm Its important to note that (I don't think) XCode isn't integrating with Claude Code, just Claude. They're using Claude's LLM, but not the agent-coding-loop infra that exists in Claude Code.
this makes me believe that using Claude Code will still be 'smarter', more useful, and more modern than whatever Apple builds into XCode.
-
@atpfm Its important to note that (I don't think) XCode isn't integrating with Claude Code, just Claude. They're using Claude's LLM, but not the agent-coding-loop infra that exists in Claude Code.
this makes me believe that using Claude Code will still be 'smarter', more useful, and more modern than whatever Apple builds into XCode.
Of course, the benefit will be that it'll have better GUI and integration with XCode itself.
I switch between Github Copilot and Claude Code in VS Code because I greatly prefer Copilot's IDE integration, but find Claude Code to be smarter, despite both using the same LLM models.
-
@secundus
I would be interested in references if you have them. I’ve been having a hell of a time finding numbers on this stuff.@niekvdpas As far as is known, AI companies do not publish detailed data, so there are more or less plausible guesstimates. Andy Masley (see link) has written several solid articles on the topic. Granted, he is just another person on the internet, but I find his argument reasonably convincing. This is not about specific numbers, but about a rough estimate of the order of magnitude of LLM energy consumption and how that compares to overall personal energy use
Using ChatGPT is not bad for the environment - a cheat sheet
The numbers clearly show this is a pointless distraction for the climate movement
(andymasley.substack.com)
-
@niekvdpas As far as is known, AI companies do not publish detailed data, so there are more or less plausible guesstimates. Andy Masley (see link) has written several solid articles on the topic. Granted, he is just another person on the internet, but I find his argument reasonably convincing. This is not about specific numbers, but about a rough estimate of the order of magnitude of LLM energy consumption and how that compares to overall personal energy use
Using ChatGPT is not bad for the environment - a cheat sheet
The numbers clearly show this is a pointless distraction for the climate movement
(andymasley.substack.com)
@niekvdpas To give an even vaguer answer. A fairly limited LLM can be run reasonably well on my four-year-old laptop. A much more capable model can be run on a machine, such as a Mac Studio. This does consume energy, but not very much. Especially given that typical LLM use does not imply sustained full-load operation.
Compared to these examples, optimized hardware or software, scale, better utilization, etc. will be more efficient.
-
@atpfm Its important to note that (I don't think) XCode isn't integrating with Claude Code, just Claude. They're using Claude's LLM, but not the agent-coding-loop infra that exists in Claude Code.
this makes me believe that using Claude Code will still be 'smarter', more useful, and more modern than whatever Apple builds into XCode.
@joshhunt Are you sure? I see two possible choices for Claude integration in Xcode ("Claude Agent" and "Claude"), and I'm not sure what they mean.
-
@joshhunt Are you sure? I see two possible choices for Claude integration in Xcode ("Claude Agent" and "Claude"), and I'm not sure what they mean.
@siracusa hmm no, I’m not sure at all. I’m not sure why they would have the distinction between the two, but I think it’s telling they never refer to it as Claude Code.
-
@siracusa hmm no, I’m not sure at all. I’m not sure why they would have the distinction between the two, but I think it’s telling they never refer to it as Claude Code.
@siracusa https://www.anthropic.com/news/apple-xcode-claude-agent-sdk
"Xcode 26.3 introduces a native integration with the Claude Agent SDK, the same underlying harness that powers Claude Code. Developers get the full power of Claude Code directly in Xcode"
"Developers using Claude Code can integrate with Xcode over MCP and capture visual Previews without leaving the CLI."
Agent SDK overview
Build production AI agents with Claude Code as a library
Claude API Docs (platform.claude.com)
"The Claude Code SDK has been renamed to the Claude Agent SDK"
huh!
-
677: I Accept the Battery Cost
https://atp.fm/677If you really don’t like AI, we have some bad news for you.
@atpfm Re the aftershow: this Yale working paper suggests that Elon's politics have cost Tesla more than 1 million sales: https://www.nber.org/papers/w34413
-
677: I Accept the Battery Cost
https://atp.fm/677If you really don’t like AI, we have some bad news for you.
@atpfm I heard your thoughts about Apple’s AI pin recording meetings and instantly thought of corporate IT. No way in hell are they allowing that without enterprise control and we all know how well Apple does enterprise tools. Just see all the consent messages from Zoom and Teams for recording to follow recording laws. And without the workplace use case, it seems that tech companies want and need the pin more than users do, just one more avenue to slop up data.
