Iv

BAR (https://www.beyondallreason.info/) is an open source TA clone that proposes massive scale and that is totally open. After the recent disappointment over a series of failures to bring a new big RTS game in the last 2 years, the RTS community talks a lot about this one.

The mechanics are old school, as with basically all RTS, but the openness allows for far more experiments than one would assume in a proprietary game.

Well, HTML was supposed to be a generic language to describe typical documents. Most websites don't need more than the default elements.

From an outside perspective, it is perplexing to see the constant back and forth webdevs do between making website more complex and rediscovering the simpler first principles

"Starting from a single base LLM"

Ok, zero data, except the data used in the teacher model.

*methane and ethane lakes

Thought it could be a useful precision.

Old people in Japan went through the deprivation of post-war Japan. Not all boomers were raised the same. It is harder for someone who has known famine to accept it is ok to throw food away for fun.

US grandparents think you can buy a house with a part time job, Japanese ones think that you could save a life with a watermelon. Different delusions.

I am also a fan of RTE (French electricity distribution network) live website: https://www.rte-france.com/eco2mix/la-production-delectricit...

We "chouettistes" are thirsty for more as well. There is almost a cult growing out of some hypothesis, we are waiting for the confirmed solutions a bit like the second coming.

factiverse.ai states GPT-4 is open source, that there is a colony on Mars and is mixed on whether Finland is a real country.

Still needs work

... daleks

> and now they managed to shake up their user's trust in leadership stability

Do users care about that? I care about features stability and avoidance of shitification.

That's why I am usually preferring open models to depending on OpenAI's API. This drama has me curious about the outcome and if it leads to more openness from OpenAI, it may gain me back as a user.

In that case, Microsoft

OpenAI switching back to being open would be one of the best news of the decade!

Either that or he refused to do something that would bring a quick money grab. 50/50 as far as I'm concerned.

The board discovered that the process `GPT5-training` that has been running for months on their über-datacenter was actually mining bitcoins.

The thing that makes Wikipedia great is that it is not a profit seeking-entity. It made it avoid all the incentives for shitification.

We need efficiency when we want to maximize comfort and minimize labor.

But nothing forbids people to pursue less efficient endeavors during their free time. There are people maintaining old cars and locomotive. There are people gardening or woodworking "inefficiently" for their own pleasure.

What we remove is the need to force people to work on these fields. Whether we abandon them altogether depends solely on our culture.

What has gone wrong is that AI researchers and futurologists have warned for decades that AIs replacing jobs was coming (though everybody was taken by surprise that artists would be first) and that we needed to change our paradigms about labor quickly.

We should be embarrassing the weight of labor lifted from our shoulders and organize society so that it becomes desirable. Replaced professionals should be allowed to go into retirement directly and access some form of basic income. There should be incentives to automating your own work.

Now incentives are on the other way: everyone tries (understandably) to make it as hard as possible to be replaced but this just leads to a world of bullshit jobs pretending to be useful while the AIs do all the heavy lifting.

"With further development, we expect"

These are the downsides. These are promises. Not actual performances.

The important aspect is not the language, but the abstraction level at which we will be.

I think you underestimate the ease with which we will be able to refactor code.

That sounds like a "there are too many competing standards, let's make standard to unify them!" situation.

I remember a pretty old interview with Linus Torvalds where they are talking about object oriented programming. The interviewer asked him if he expected a similar paradigm change in the coming years and I remember being surprised by his answer: (quoting from memory) No, I don't see anything big coming. Probably the next change will be caused by AI.

Yes, differentiable code is already a new paradigm (write a function with millions of parameter, a loss function that requires more craft than people realize and train). That has a property that used to be the grail of IT project management: it is a field where, when you want to improve your code performance, you can just throw more compute at it.

And I think that the clumsy but still impressive attempts at code generation hints at the possibility that yet another AI-caused paradigm change is on the horizon: coding through prompt, adding another huge step on the abstraction ladder we have been climbing.

Forget ChatGPT coding mistakes, but down the road there is a team that will manage to propose a highly abstract yet predictable code generator fueled by language models. It will change our work totally.

You can even get ChatGPT to write programs to query different APIs or to do calculations.

Had ChatGPT been an open model, like OpenAI was supposed to produce, this kind of applications would have seemed obvious and happened in the first two weeks after release.

Hell it could even be the proof of work for a usable crypto-currency. "Prove that you lowered the error rate compared to SOTA and earn 50 ponzicoins!"

About 30% of workers are in "bullshit jobs".

https://theanarchistlibrary.org/library/david-graeber-bullsh...

Here is a dystopia we don't hear about too often: instead of robotic overlords, infinite inequality or labor-less utopia, automation could instead be met with an increase of bullshit jobs validating each others. (Iain Banks' Walking On Glass is the closest I found)

Give me open source dev tools for the iPhone and I'll jump.

While it is a closed garden, I'll begrudgingly accept it can be marginally better in some fields than other options, but Apple tries very hard to be a proprietary island in a world that has switched to free software.

Usually CE regulations do not care if you hired an auditor or not. They only care about the regulations being obeyed. My understanding is that warning the users of the level of security they can expect, and handling security flaws reasonably is probably going to be enough.

CE is typically obligation of result, not obligation of effort. Do you have source on products that can't be self-assessed?

CE certification is based on results, not efforts. You don't "obtain CE certification", you simply are liable for non-compliance if your product is non compliant, even if you gave money to someone who runs a "CE certification" business.

Note that there are tons of CE certification. For radio cert for instance they don't need to open the box.

French on US intl here ä ö ü ß are easy to make on such a setup but also the accented letters like é à ô or the infamous ç. It is even easier to type them in capital than it typically is on AZERTY (French keyboard) or QWERTZ (German one)