Richard van Yperen

Now for €99,- per month you can have your Tesla drive you around.

Great news, time to switch from Claude now we can’t use that for Openclaw anymore.

@karpathy I noticed the people who really give in to their curiosity and experiment are the ones who are impressed by what we can achieve with AI and keep thinking of new ways to use it.

Judging by my tl there is a growing gap in understanding of AI capability. The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code. But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along. So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions. TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.

@rileybrown Wandering myself, also looking to see what subscription is still able to be used with Hermes/Openclaw. Dont have the budget for API but do want to keep experimenting with agents.

Is Hermes better than OpenClaw or is it yet another psyop on the timeline?

@marcvanderchijs @DavidCuriousMan I get that it might be good to be a bit scared, but a lot of people freeze when scared, in my experience if you take the time to help people experiment with AI they get excited and start thinking in possibility instead of risk.

@DavidCuriousMan @vanYperen Until now things never changed very fast, so they assume the AI changes will be slow as well. Unfortunately we know that is not the case.

Because of the exponential growth of AI capabilities and AI agents, the digital divide is growing faster than ever before. If you just use AI as a search engine, you have to spend a few hours to dive in deeper. Today, not next week or month. I think at least 70% of office jobs can now be done by agents. Maybe not 100%, but to the point that 1 person can do the work of 20 humans by managing a swarm of agents. Learn how to automate your own job before others do it for you, it’s easier than you think with tools like Claude Cowork (and that should make you scared!).

@internetarchiva One of the reasons I build KeepOfEchoes.com

For ten years, an elderly man listened to a voicemail from his late wife every single day. When it was accidentally deleted, eleven engineers spent three days working nonstop to recover it.

Not again

Really hitting those new Anthropic Claude Code limits 🤨

@Tobby_scraper I build the Home Inventory app, Spullio.com

Underrated iOS niches for vibe coders: - Car Maintenance Loggers $55k MRR - Fishing Spot Maps $35k MRR - Home Inventory Apps $45k MRR - Wine Cellar Trackers $40k MRR Small audience. Lowcompetition. Easy to rank. Check Niches Hunter for more ideas! 🚀

I will try this every time I hit my Claude rate limits. Always enjoy having multiple models discuss and work together.

@LinusEkenstam A lot of projects, but Spullio.com and KeepOfEchoes.com are live and I love them.

Can I see some links of what you’ve built with Claude code? Blow my mind

@rileybrown And doesnt really work on mobile (except single session dispatch)

There’s like no reasons to use cowork over using Claude code. It’s just less capable.

Tried and failed to reply... 🙄

@amorriscode Tried and failed

if you're bored waiting for claude to finish doing work, start another session. life is too short to be bored.

Wait, I was still waiting over here in the Netherlands 🙄

Was watching F1 and keep having token rate limits anxiety due to the battery management.

Was waiting for Perplexity Computer to come to Pro plans, guess this might be more fun

@cryptopunk7213 Wonder why transportation is zero, don't they see self driving (AI) as something real?

wow Anthropic just published a crazy report on AI replacing your job and er... you might want to look at this: - #1 most at-risk jobs are computer programmers, financial analysts (rip excel bros) and customer service - most at-risk workers are female, white, older and higher paid. - BUT high-risk jobs *aren't* firing employees... they've STOPPED HIRING. biggest victims: college graduates (4X more likely to be fucked) - entry-level hiring has dropped 14% since chatgpt launched (for highest risk jobs) - SAFEST jobs are... bartenders, dishwashers and lifeguards - any manual labour that AI can't automate (yet) this accounts for 30% of the job market. - this was the scariest part: AI models are capable of automating most work TODAY but are prevented because of law and slow company adoption. so its not even a fucking skill issue its an ADOPTION issue. - now its important to understand that the study is based on real world data but also 'theoretical' intelligence. so take it with a pinch of salt. some jobs (manual labor) didn't even meet min. data reqs i applaud anthropic on being so damn transparent - they're literally the company behind claude who will be responsible for these impacts studies like this will help us figure it the hell out. LOT of change coming this year.