fsndz

> AI-generated text has none of that effect on me. My brain just shuts down as soon as it detects it. My eyes glaze over and my fickle attention flickers away.

Does this apply to all AI-generated text? Don't you use ChatGPT or Claude, and thus read AI-generated text?

A few days ago, I asked myself this question: “What happens when AI can say what you want to say better than you?” That question only scratches the surface of the malaise. For more than a year now, I have struggled with what to do, as someone who likes to write, when writing becomes ‘cheap’. Or should I say, to be more precise, that it is actually generating written artifacts that has become easier. There is a feeling that, now that one can write enormous amounts of blogs, essays, and reports, each one becomes less valuable. Text is no longer rare and precious like in ancient times. Just as synthetic diamonds are killing the natural diamond market, there is a sense that the synthetic writings we currently have wide access to are degrading the value of traditional human writing.

this is impressive and actually matches my expectations in terms of near term AI progress. we are going to continue to seem impressive progress in coding & related, anything where verifiability is scalable in an automated way: https://transitions.substack.com/p/a-quantum-of-ai-progress?...

expertise at navigating accidental complexity can easily be mistaken for engineering expertise.

we can't automate it anyway and vibe coding is overrated: https://medium.com/thoughts-on-machine-learning/vibe-coding-...

I did that, burned 2.6B tokens in the process and learned a lot: https://transitions.substack.com/p/what-burning-26-billion-p...

As part of the Holistic Agent Leaderboard (HAL) initiative at Princeton CITP, we evaluated more than 220 agent runs across 9 benchmarks, the equivalent of over 20,000 agent rollouts across 9 models and 9 benchmarks for a total cost of $40,000. The benchmarks are: AssistantBench, CORE-Bench Hard, GAIA, Online Mind2Web, Scicode, ScienceAgentBench, SWE-bench Verified Mini, TAU-bench Airline, and USACO.

In that process, we “burned” 2.6 billion prompt tokens and learned a lot along the way. In this article, I’d like to share some of the insights we gained, with a particular focus on the GAIA benchmark.

By that definition, the ChatGPT app is now an AI agent. When you use ChatGPT nowadays, you can select different models and complement these models with tools like web search and image creation. It’s no longer a simple text-in / text-out interface. It looks like it is still that, but deep down, it is something new: it is agentic… https://medium.com/thoughts-on-machine-learning/building-ai-...

and people are still saying vibe coding is overrated? nonsense: https://www.lycee.ai/blog/why-vibe-coding-is-overrated

Exactly. I think the study is a good reminder that we really have to be careful about the productivity gains attributed to AI. Main takeaway imo, despite limitations from the study, is AI is not a panacea, it can increase productivity, but only if used 'well' and with the good workflows in place, and in the right context.

I mean, hacker news is still the same aren't they using AI to completely make this website more of whatever it was before ????

exactly !

Klarna tried this narrative shopping strategy first and it backfired: https://fsndzomga.medium.com/i-have-no-confidence-in-klarna-...

I am tired of seeing this. First the Klarna dude did it and it backfired. Now Andy... People fail to grasp the fact that building AI agents is no longer enough, you need to do more: https://medium.com/thoughts-on-machine-learning/building-ai-...

I now understand why some people say MCP is mostly bullshit + a huge security risk: https://www.lycee.ai/blog/why-mcp-is-mostly-bullshit

It's not. can we stop doing this ?

the smolagents library is also pretty nice to do the scaffolding around the model. Text to sql seems simple in demos, but to make it work in real life complex cases is very hard: https://medium.com/thoughts-on-machine-learning/build-a-text...

xarray is nice

I can be frustrating at times. but my experience is the more you try the better you become at knowing what to ask and to expect. But I guess you understand now why some people say vibe coding is a bit overrated: https://www.lycee.ai/blog/why-vibe-coding-is-overrated

Never bet against science !

it's always opt-in until it isn't

and setting up postgresql on a simple VPS is so easy... You can literally ask Gemini 2.5 Pro or o3 or Sonnet 3.7 and do it in 15-30 minutes... Learned helplessness is really something and vibe coding is overrated imo: https://www.lycee.ai/blog/why-vibe-coding-is-overrated

why is Google more evil than say OpenAI ?

More and more people are coming to the realisation that Google is actually winning at the model level right now.

True.

Deep learning models will continue to improve as we feed them more data and use more compute, but they will still fail at even very simple tasks as long as the input data are outside their training distribution. The numerous examples of ChatGPT (even the latest, most powerful versions) failing at basic questions or tasks illustrate this well. Learning from data is not enough; there is a need for the kind of system-two thinking we humans develop as we grow. It is difficult to see how deep learning and backpropagation alone will help us model that. https://medium.com/thoughts-on-machine-learning/why-sam-altm...

why would anyone accept to expose sensitive data so easily with MCP ? also MCP does not make AI agents more reliable, it just gives them access to more tools, which can decrease reliability in some cases:https://medium.com/thoughts-on-machine-learning/mcp-is-mostl...

true. SAP is more of an entrenched SaaS. Once you have it, the switching costs are too high later on. What an amazing kind of biz. The stickiness is crazy.

okay give me the link to a LLM-based system that does not hallucinate then

MCP is basically what RAG was 2 years ago.