uxhacker

Basically the article is arguing that distillation is a moat leak from USA to China.

But what surprises me is that the frontier labs can’t detect the Chinese making the distillation attempts?

It’s not just that. Working with multiple agents and tasks switching will increase cognitive load significantly leading to both poor decision making and increased stress.

https://pmc.ncbi.nlm.nih.gov/articles/PMC7075496/ https://pmc.ncbi.nlm.nih.gov/articles/PMC7614709/

In a car with Recirculation Mode on Levels routinely spike between 1,500 and 4,000 ppm within 20 to 30 minutes.

I wonder how many driving accidents can be saved by having a co2 monitor in the car.

So in Poland with Starlink I easily get 120 mbp/s but often in Kenya which I visit the maximum speed is just 10 mbp/s. Often the local 5g network is faster. The reason I believe that this is the case is dues to congestion on the Starlink network in Kenya. Ie. too many users in Kenya.

The article gives valid advice in using AI on tasks AI is good at, but the article could be clearer on is what AI is good at and what AI is not so good at.

Actually I am been asked to pay for Reuters

Edited : this looks like it is in paywalled https://uk.news.yahoo.com/austria-lobbies-eu-host-anthropic-...

For a bit more context the Belarusian activist built on the anti communist Polish Activist Waldemar "Major" Fydrych who in the 1980’s was arrested by the communist authorities in Poland for handing out female sanitary products.

As he said “The Western World will find out much more about the situation in Poland from hearing that I was put to jail for giving tampons to a woman, than from reading the books and articles written by other people from the opposition.”

A counter point to Paul Graham’s How to Earn a Billion Dollars speech at the Oxford Union. https://paulgraham.com/earn.html

I agree more with Graham, but it is worth a debate as so much power is tied up with People like Elon Musk.

What is surprising is that GSM-R is 2g. Does not 2g have many security issues?

Their argument on page 10 is that now agentic coding reduces the effort of writing code there will be far more failures unless you validate the idea properly.

We are actually seeing that in that the number of apps on the app stores is increasing but usage is not increasing.

Some would argue that the right process will lead to the right results.

Your making an argumentum ad verecundiam which even if you are right means we have to discredit it.

It’s poor science to make an argument on authority, if you know the science then you should be quoting the published research and not relying on others so called expertise.

Groupthink is very much the scientific method. According to Imre Lakatos the key question is does the group expand knowledge or contract it (very rushed reply as about to catch a flight)

I wonder how much of this is to lower the valuation of Anthropic.

Reuters story covering the story : https://www.reuters.com/technology/openai-considers-drastic-...

Without a paywall

Also according to the Financial Times that this confidential filling gives employees who are considering to sell shares transparency.

According to this from oracle the technology is there for closed loop systems. https://www.oracle.com/news/announcement/blog/closed-loop-co... But is it green washings?

I wonder how many people saying similar things have the same background, but having said that she was VP of Engineering at GitHub, and VP of Engineering at Heptio.

Should we not be counting function points rather than LOC’s.

Lines of Code is a meaningless measure. It should also be easy to count function points using AI.

Often to me it seams like using MA is like letting a million monkeys lose.

Has ai forgotten about high level design? Surely all it needs to know is what the methods, objects or functions in the code base actually does and the actual code it is meant to be fixing?

I wonder if half the issues is that the LLM try to change too much?

So what harness are you using? And what LLM’s

I think some justify it as SpaceX plans to offer hosting in space, and then use Starlink to distribute it.

The argument is dialectical. In other words both sides are right.

There is a really important question that is been lost between the anti and pro AI camps which is really answering what AI is good at and what AI is bad at, and what is the root cause of the weaknesses. Is it intrinsic to the models that LLM's use, or the way thy have been trained. In this knowledge is where the gold mine is for the next start up.

For example AI is very good at answering well defined questions, but suffers from premature closure. It will not know if it has all the information to answer the question. So whilst AI will score better than a doctor or a lawyer on a domain question it will not necessarily gather all the evidence needed to answer the question properly. Knowing this whilst using a LLM is a super power.

There is also a large gap Usability issue in that often the LLM does not really know the humans context, in other words a context collapse. It does not know if you specilised in your domain or just asking for fun.

We should be exploring and debating these gaps.

Try this with ChatGPT or GROK or Claude

How many days of the week contain the letter d?

The answer I get with ChatGPT, and Grok is 3 and 6 with Claude.

Yes, I agree. There are two theories here Quantity has gone up but quality has gone down, or people need less apps as the LLM answers their needs.

Trying to answer this is important to try to answer the future. What startup to build

A really interesting point in the article that even though the number of IOS apps Has increased by 80% the number of reviews has actually decreased.

The number Apps with significant usage has gone down slightly.

Is this suggesting less critical thinking?

I think this article echos a lot of what people have been saying here on HN, that there is a massive funnel of productivity.

The article says that whilst the number of edits to individual files has increased by 300% actual releases have only increased by 30%.

I even find this 30% seams higher than what I am seeing.

I love the quote by the Japanese engineer Ohno that a broken machine or system makes a lot of noise. I think it sums up the current situation precisely.

And archived https://archive.is/5nHAP

So how much more improvements are there for efficiency in the Claude code base if they are using react for a tui, in the rest of the code?

I also wonder about the wasted cycles and just the environmental damage caused by all these wasted cpu time . (Edited added a comma for clarity)

Why did the account recovery system need AI. Surely just an email would do? What added value would AI add?

What I can’t understand is how she was able to publish the book, but is not able speak publicly about what happened.