Hacker News new | past | comments | ask | show | jobs | submit | nsoonhui's comments login

  We will also begin deprecating GPT‑4.5 Preview in the API, as GPT‑4.1 offers improved or similar performance on many key capabilities at much lower cost and latency. GPT‑4.5 Preview will be turned off in three months
Here's something I just don't understand, how can ChatGPT 4.5 be worse than 4.1? Or the only thing bad is that the OpenAI naming ability?

They tried something and it didn't work well. Branching paths of experimentation is not compatible with number-goes-up versioning.

No, building bridges is a lot simpler because we have a set of coding practices to follow ( eurocode, ACI, British standards and whatnot). The performance of the buildings follow the deterministic Newton's laws, and you even have software packages written for that.

Your clients don't change the specs half way through and expect you to provide an accurate estimate and at the same time, don't intend to pay you extra. You can forecast the cost of building to a very accurate degree because all bridges are more of the same, unlike software which by definition,is new every time because each time the requirement is different.

And so on.

I know because I work in civil engineering software field.


The comments from HNers in this case are interesting.

Usually if a teenage hacker builds something, the HNers would respond with enthusiasm, but then, this is a guy who builds *something* and *actually* makes a good business out of it, at the same time maintains his high school life, and all we have is skepticism and discouragement.


It's almost as if people are reading the article and critically evaluating the product.

Fwiw scams can be inspiring too if the sense of humour is evident, and/or the target of the scam are bad faith actors. (E.g.their earlier app "totally science") Blame the reporters for simply saying "it's ironic"?

It is _not_ a good business; the thing does not, and cannot, work.

It's good business, just bad product.

Interesting take, care to explain more exactly how it is much better?


It's exactly "much" better!


Any circumstances change that prompted Alphabet and Wiz to redo the deal?

I am not aware of what are the exact roadblocks that stopped the deal last year, and it isn't clear from the article that they have been removed.

23 billion is already a huge sum, so more money than that isn't likely to cause a change of heart, I guess.


You wonder if Wiz is having second thoughts about IPO just as we go into recession


What recession?


> What recession?

There are increasing signals of an upcoming recession, but with Trump's unpredictability there is a lot of uncertainty...

https://www.nbcnews.com/data-graphics/recession-warning-sign...


An article like this gets published a few times a year, every year in the last 50 years. Some years it comes true!


10 Billion Dollars / 50% more value for the employees stock is a big deal though. I know this might be unheard of but I think the CEO understood the value and wanted to end up with a deal that made even later hires a life changing sum

There were a lot of haters on HN and elsewhere when he turned down $22B, let’s see others demand 50% higher valuation in front of the alphabet board and actually get it. He literally just believed in the product

I mean even now this thread is slowly filling up with conspiracy theories on how they agreed to 50% MORE money only because they’re about to crash. Come on…

They wanted a respectful offer from Google, they didn’t get one, now they got one. The fact the companies kept talking after the last round of press shows that there is substance in the product

Imagine you just smeared Google in the global media, and they call you the next day trying to get back together


Maybe anti-trust concerns plus a massively more amenable administration and DOJ? Biden + big tech was always an uneasy marriage at best.


Zvi posted a response[0] to this post, with the following comment:

  Ignore the first sentence in Tyler Cowen’s post here, where he asserts that Manus is ‘for real, and ahead of its American counterparts.’ That’s a rather silly way of summarizing the situation, given everything we now know.
[0]: https://thezvi.substack.com/p/the-manus-marketing-madness


I suppose that you are using a dynamic language? Static typed languages have less of this problem.


This is a hypothetical "I". Personally I am deeply passionate about delivering shareholder value, producing high-quality code, enthusing stakeholders, tabs vs spaces, and so forth...


Blog post from Scott Aaronson https://scottaaronson.blog/?p=8680


>> It is not a compiler error. It is never a compiler error (2017)

No, not always true. Even in modern compilers -- as matured and as modern as VS 2022-- you would still get bug.

I found one[0]. In my case it's easy to tell it's a compiler bug because the program just can't compile properly. But it's also not easy to reproduce, which just proves how well tested compilers usually are.

0: https://github.com/dotnet/roslyn/issues/74872


>> Amodei's / Hassabis' comments in particular came off as so arrogant and annoying.

Exactly which part of their writings comes off as arrogant to you? The only point in Amodei's article[0] that could be remotely be interpreted as arrogant is this:

  All of this is to say that DeepSeek-V3 is not a unique breakthrough or something that fundamentally changes the economics of LLM’s; it’s an expected point on an ongoing cost reduction curve. What’s different this time is that the company that was first to demonstrate the expected cost reductions was Chinese.
Maybe I'm different, but it really does sound reasonable judgement to me.

[0]: https://darioamodei.com/on-deepseek-and-export-controls#deep...


"Humanity was bound to develop anti-gravity engines eventually, it just so happens that the country that did that was Rwanda."


China is the second richest country in the world, and the one with the most computer scientists. Americans sometimes think the rest of the world is far behind but none would compare China to Rwanda in AI.


Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: