Grok is quite good for explaining tweets, summarizing meme videos and chatting about celebrities. But chat and image inference only, it was performing really poor in agentic work the last time I tried.
I made something similar to this project, and tested it against a few 3B and 8B models (Qwen and Ministral, both the instruction and the reasoning variants). I was pleasantly surprised by how fast and accurate these small models have become. I can ask it things like "check out this repo and build it", and with a Ralph strategy eventually it will succeed, despite the small context size.
I tried to sign up for Gemini this weekend but gave up after an hour. I got stuck comparing their offerings, looking for product pages, proper signup, etc. Their product offering and naming is just a mess. Cloud console. AI studio, I was completely lost at some point.
$20 Google AI Pro and Google Antigravity IDE, which gives you access to Claude Code, is a pretty decent offering for agent coding. On top of that, NotebookLM and Google Labs has some fun tools to play with.
I love these thought experiments. Looking at the code size, it would have been possible for someone to come up with this back in the days, similar to the idea of a million monkeys on a typewriter eventually producing Shakespeare.
I actually get more mileage out of Claude using a Github Copilot subscription. The regular Claude Pro will give me an hour or up to 90 minutes max, before it reaches the cap. The Github version has a monthly limit for the Claude requests (100 "premium requests") which I find much easier to manage. I was about to switch to the max plan but this setup (both Claude pro and Github Copilot, costing 30 a month together) was just enough for my needs. With a bonus that I can try some of the other model offerings as well.
In practice, how does switching between Claude and GitHub Copilot work?
1. Do you start off using the Claude Code CLI, then when you hit limits, you switch to the GitHub Copilot CLI to finish whatever it is you are working on?
2. Or, you spend most of your time inside VSCode so the model switching happens inside an IDE?
3. Or, you are more of a strict browser-only user, like antirez :)?
I always start in the Claude CLI. Once I hit the token limit, I can do two things: either use Copilot Claude to finish the job, or pick up something completely different, and let the other task wait until the token limit resets. Most importantly, I'm never blocked waiting for the cap.
Good to hear that’s working. When I was using copilot before Opus 4.5 came out I found it didn’t perform as well as Claude Code but maybe it works better now with 4.5 and the latest improvements to VSCode. I’ll have to try it again.
So you buy a new TV, unpack and install it, and then when the whole family is gathered around, you suddenly get this confirmation on the TV if you agree with their T&C. Are you supposed to reject them and return the TV at this point? T&C should be part of the purchase agreement, instead of being forced upon the user while using the product after purchase. Any one-sided change of T&C after purchase should be invalid and punishable.
Not everyone feels like that. Yesterday the app of my tv provider on my Samsung TV home screen suddenly shows a Prime icon in its place, prompting to install the app if you use muscle memory to control the TV. I am unable to remove this annoying ad. I really really hate ads and will go to great lengths to avoid seeing any in my private home. So I see this as an invasion of my privacy. Not buying Samsung anymore.
I've been thinking about this as well - make a small device that in real time detects ads and turns off audio an video while it's playing. I'd rather see a blank screen than an ad. That way, the whole ad pyramid scheme stays intact while the conversion rates plummet.
Isn't the segment who will set this up also likely to have a low conversion rate to begin with?
You'd need to make it so easy that it becomes fully mainstream. I suspect that's what happened to adblockers, it got a bit too "standard" for (Google's) comfort.
Same here. I've done this for podcasts (not in real time) and it works great. TV should be easier in some ways since the video stream and captions can also indicate an ad.
I used to find when listening to a good many podcasts with VLC there would be:
> ... See you after the break.
brief pause
> And we're back ...
Unfortunately, most ads are now burnt in. The 10 second advance will skip through them, but as it's usually the host parroting the ad text and it's easy to over shoot.
reply