HN user

fedeb95

547 karma
Posts21
Comments538
View on HN
news.ycombinator.com 3y ago

Ask HN: Research on GPT3 generated text recognition?

fedeb95
1pts0
news.ycombinator.com 3y ago

Ask HN: Interest in FOSS tool to analyse market data?

fedeb95
1pts0
news.ycombinator.com 3y ago

Ask HN: How to handle financial calculations precisely?

fedeb95
18pts17
www.scriptkid.it 3y ago

Almost Patterns

fedeb95
1pts0
news.ycombinator.com 3y ago

Ask HN: Is there academic research on software fragility?

fedeb95
46pts46
news.ycombinator.com 3y ago

Ask HN: Which is the last stock you bought and why?

fedeb95
6pts8
www.scriptkid.it 3y ago

Information aggregators: why HN is better

fedeb95
7pts1
news.ycombinator.com 3y ago

Ask HN: Why reusing e2e tests output is bad?

fedeb95
1pts4
news.ycombinator.com 3y ago

Ask HN: Has someone already trained some AI on sacred-texts.com?

fedeb95
7pts1
news.ycombinator.com 3y ago

Tell HN: GitHub Copilot scraping OS projects is good

fedeb95
7pts8
news.ycombinator.com 3y ago

Ask HN: Past HN post of a website designed like a garden?

fedeb95
1pts3
news.ycombinator.com 3y ago

Ask HN: Am I the only one not caring about GitHub Copilot?

fedeb95
52pts59
news.ycombinator.com 3y ago

Ask HN: Does booking flights from a different country give different prices?

fedeb95
4pts3
news.ycombinator.com 3y ago

Ask HN: Source code (Java) parser and/or static analysis tool

fedeb95
2pts2
www.scriptkid.it 3y ago

Probability of Luck

fedeb95
1pts0
news.ycombinator.com 4y ago

Ask HN: Why the EU did a law only for phone chargers and not, say, laptops?

fedeb95
1pts4
news.ycombinator.com 4y ago

Ask HN: Is there a news feed like HN but less computer science/science focused?

fedeb95
3pts2
noblogo.org 4y ago

Hurst exponent in crypto (4 years later)

fedeb95
2pts0
noblogo.org 4y ago

A parallelism between CTF challenges and chess

fedeb95
1pts0
github.com 4y ago

Show HN: Netplot, plot network connections in the terminal

fedeb95
8pts0
github.com 4y ago

Sinfract

fedeb95
1pts1

the point I see is that even if you automate all that is possible to automate, there are some core activities that can't be delegated to machines in order for a product to feel very good. May it be the design, or whatever: I don't have an opinion on which aspect must be curated by a human, but some aspect has to. Maybe this aspect is simply AI review and prompting, or AI infrastructure creation and maintenance... I don't know. But the point is that this part of production exists, and it's what makes a useful product something you want to buy more than another useful product.

this is about open hardware.

You really don't need to build a Kobo clone. By being open, you can fit your needs. If your needs are already covered by Kobo, that's fine.

I can't believe someone shares my point of view. We have already many non humanoid robots to do boring stuff (underwater maintenance, stuff like that). For some reasons people tend to think that for robots to do stuff close to humans, we need human like robots.

We do not. I want a swarm of insects-like robots to fight mosquitos. A snake one that fixes my pipe. A spider-like robot that cleans surfaces. Etc.

I use AI a lot but am not addicted to it. I know it seems something every addict would say, but simply it makes things faster. You can't really take responsibility for something it makes that you don't already know how it operates, or how should operate. That's the line between what you can sell to your client, or your boss, without it blowing up in your face when things go wrong. In theory everything I build with AI I can build by hand, but in many more days.

A builder is not addicted to their tools, but he won't certainly start a project without them, if they are available. Yet, he could work without some of them. Give a circular saw to someone that can't use it, and nothing good comes out of it. Give AI to a non coder, and nothing good (that lasts) will come out of it.

That words are a byproduct of consciousness is a bold claim and requires some justification. I'm not implying it's backwards also for humans, but that the statement might not be true, and something else could be. Also, defining consciousness is hard.

For instance, you can dream words.

I think most people forget the graph-like nature of scientific research. You don't have n useful papers and m useless ones by themselves, you have an interconnection of those. There may be isolated cliques of uselessness, but there isn't a clear correlation between academia and private research.

Many ideas come from philosophy, which many find useless.

Heraclitus discovered change back in ancient Greek, I don't know where we would be in scientific research without that (deliberately ignoring the debate about the originality of what we know about Heraclitus work). I bet his contemporaries found his "research" useless.

Not entirely true. Many science is gatekeeped, as well as other types of information. May books require illegal services to be obtained, or money (when available). Information about facts is buried in a lot of misinformation. Free flow is very hard to obtain!

the real money is in the coding surrounding models to make them efficient at specialized tasks. Casual users want general purpose models, and AI chat apps will stay for them. Most programs can benefit from a specialized AI that can be local, and #programs >> #users.

moving around could be accomplished by traveling, in decreasing order by efficiency (by those who can, obviously):

1) cycling 2) walking 3) train 4) cars 5) airplanes

and the frequency could follow an exponential distribution.