News for hackers. Stuff that matters.

Mmskynettio

Re:We went solar and here are the real numbers (2021) on Sunday June 09, 2024 @04:06AM (#40622790)
I have a balcony PV thing. It was easy to install and saves money. Will recupe it in a few years.
No why? Are you not able to read arguments and argue against them instead of just 'whatever your answer is'?
But but but Crypto will change the world?! And everyone should have their private wallet? And who cares about recovering your funds because everyone of us will be handling private data / secrets (like passwords or keys) perfectly, always!111

Cryptobros telling you never to use an exchange due to FTX and other examples, also its super easy to use...

What stupidity

I do and i found plenty of issues while doing code review.

But i started to read a lot more code than what i did 10 years ago

But thats the point. You can use OpenAI through Azure and other means. If you trust MS (who has basically data from millions of companies) why would it not work out with OpenAI usage?
So you are not using Office 365?

Because our company does.

I'm getting jun devs who are brainstorming and troubleshooting with chatgpt before they start talking to me.

I have a good intern who is much faster with chatgpt than before and learning well.

The problem is that we don't know how we do reasoning.

I calculate stuff by following a formular after i pattern detected a problem i already know.

Plenty of humans are not able to solve those math problems.

If the future of llm / ai becomes a LLM with multi modal and mixture of experts and that solves those reasoning problems, we still don't know if this is a different type of reasoning than what humans do.

Does it need to do reasoning perfectly?

If it created meta concepts from billion words on the internet and has meta models which are correct and are more and better than an avg human, isn't it actually good in reasoning?

Its a very narrow thing to say 'is that so many think its actually reasoning' to say AI is just hype or everything we are doing is a waste etc.

There are human benchmarks they are winning at. The critic could be more that we don't have enough benchmarks.

Its not an AI hype. A hype is defined as something which gets oversold: "promote or publicize (a product or idea) intensively, often exaggerating its benefits."

Just yesterday I visited a google cloud summit and one person from bosch told the audiance how they are now able to work with less external agencies like texting, graphicsdesigner and photographers for their materials.

It already saves money, has real impacts and continues to progress.

We are also don't know what ChatGPT 5 will bring, because they say this will do more reasoning than before, but we already are working (people/our socity) on solving this in different ways: From code which creates a unit test first and than the code, to different type of architectures.

For me, 2024 was the LLM cost reduction year and the LLM gets a big context window year.

AI doesn't need to be ready tomorrow, but its capabilities are already really good. And i know plenty of people around me who are a lot less interesting to talk to than any llm (from a human skill/knowledge point of view).

llama 3 was also a big achievement 2024. Facebook shows that better data leads to better quality for smaller models.

We haven't not only entered the AI ara but also the 'gather all the knowledge we can, quality check it and refine it because now we can actually do something with it' ara.

We are in the feedbackloop knowledge ara.