Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

DeepSeek V4 Flash 0731

arcprize.org

Post reply on HN
Print view View original

441–450 of 474 posts

Re: DeepSeek V4 Flash 0731

#441

Post by F7F7F7 »

solenoid0937 wrote: post #372
AussieWog93 wrote: post #290

Earlier quoted context omitted.

To put actual numbers on it, since using AI to start solving all kinds of bottlenecks/inefficiencies in our small business, we've seen monthly net profit go up by around $4,000 USD. These are semi-permanent fixes, and the tech is only partially deployed. I am the only one using it, and I only use it part time. We've just spun up our first Hermes agent, with direct API access to our main inventory system and that's ex…

I think HN doesn't really understand fixed costs, I spend a few hundred dollars per day on Fable and the costs are irrelevant compared to what we make.

I've never understood this line of thinking...

Are you saying we shouldn't care about the future of affordability and access because at this moment we have seemingly endless access?

Sounds extremely short sighted.

Re: DeepSeek V4 Flash 0731

#442

Post by don_esteban »

petervuto wrote: post #400
RealWed5 wrote: post #313

Earlier quoted context omitted.

"оur first Hermes agent, with direct API access to our main inventory system" – let me assure you that absolutely nothing can go wrong here, mate. /s

When you do things in life, sometimes things go wrong. Oh no! This is such a ridiculous objection for how beloved it is. Wide swathes of the public can't cope with any adversity or risk.

No, the point is that with human in the loop the downside is (usually!) rather limited, as common sense would stop obvious fuckups (ok, not always, but still).

With an agent (especially incompetently employed), the danger of unwittingly destroying your company (or at least, the crucial data/reputation) is rather higher. We are notoriously bad at estimating the downside risks in complex systems.

The most obvious case is the downside risks in complex financial constructs... things look great for a while ... until a sudden surprising collapse arrives and totally destroys all the upside you think you have created.

Re: DeepSeek V4 Flash 0731

#443

Post by Kuyawa »

I use DeepSeek on a daily basis and I didn't spend $10 in the whole month of July delivering eight fully functional apps.

Btw if you need an app I may deliver it to you in ten minutes for just five cents if I'm in the mood. Just let me know.

Re: DeepSeek V4 Flash 0731

#444

Post by paulddraper »

Kuyawa wrote: post #443

I use DeepSeek on a daily basis and I didn't spend $10 in the whole month of July delivering eight fully functional apps. Btw if you need an app I may deliver it to you in ten minutes for just five cents if I'm in the mood. Just let me know.

I assume you got the email about them increasing prices "significantly" soon (but no hint on what significantly means).

Re: DeepSeek V4 Flash 0731

#445

Post by canadiantim »

dd8601fn wrote: post #325
FernandoTN wrote: post #194

Earlier quoted context omitted.

I think you're overlooking the fact that for long-horizon tasks, even small errors compound over time and can lead to catastrophic outcomes. For simple queries, we have reached the threshold since the beginning of the year, and models are good enough from every provider to make a meaningful difference between one another. (ChatGPT, Claude, Gemini, Grok, MuseSpark, Kimi, DeepSeek, GLM...) The real unlock will be, and…

I have a silly (but honest) question. What's an example or two of a > 24hr task that people are actually asking something to do? Like real life ones.

Working through the total backlog of issues in a repo that may have accumulated from planning sessions

Re: DeepSeek V4 Flash 0731

#446

Post by Kuyawa »

paulddraper wrote: post #444
Kuyawa wrote: post #443

I use DeepSeek on a daily basis and I didn't spend $10 in the whole month of July delivering eight fully functional apps. Btw if you need an app I may deliver it to you in ten minutes for just five cents if I'm in the mood. Just let me know.

I assume you got the email about them increasing prices "significantly" soon (but no hint on what significantly means).

Yep I got the email, but I am not worried. Building an app for 5 cents will cost me what 10 cents? A full dollar? Still a bargain. Still giving me two months and 29 days free for something that took me three months to do

Re: DeepSeek V4 Flash 0731

#447

Post by filterfish »

WarmWash wrote: post #337
jacquesm wrote: post #317

Earlier quoted context omitted.

They can do the same thing in the US. What are you going to do, sue OpenAI or Anthropic?

Well, yes, because the courts and the AI labs are totally separate entities. If they are doing it to you, they are probably doing it to others, which makes an easy class action

Is any class action lawsuit "easy"?

Re: DeepSeek V4 Flash 0731

#448

Post by WarmWash »

DiogenesKynikos wrote: post #428
WarmWash wrote: post #426

Earlier quoted context omitted.

China has laws that serve the state, not the individual. They are there for the party to use to prosecute. So you could file, but considering the state is the one responsible for holding the data, and the one that controls the courts, it's not going to go anywhere. Remember China is still an authoritarian dictatorship. One leader with absolute power for life. Don't let the facade misguide you.

The case mentioned in the article I linked to involved a private citizen suing a tech company, based on China's GDPR-like laws. He won.

Sure, because the state didn't have a stake.

Trying suing when it's in the states interest, like it is to build AI datasets on western data. For reference, no case in China has ever been ruled against the government. They don't have things like judges smacking down executive orders or refunding tariffs.

Re: DeepSeek V4 Flash 0731

#449

Post by Art9681 »

abixb wrote: post #162
LaurensBER wrote: post #9

I've been using it extensively since the release and the best summary I can give is that it's good enough to use it for (almost) everything and cheap enough that the cost are irrelevant. I'm running it in Oh My Pi with a second instance running as "advisor" and even with 5-6 active sessions (effectively 12 streams) I'm struggling to spend more than 5 bucks per day. OpenCode Go even has double limits temporarily so fo…

If what you're saying is true and accurate, then US-based AI labs are in big trouble. The only saving grace might be some sort of a 'national security' proclamation banning the use of state-of-the-art Chinese (and non-US) models across US federal and state governments and large enterprises (especially ones with federal government contracts), but even still, US AI labs will probably lose out massively on international…

It's not necessarily true. OP has not provided actual objective figures. The nuance is in tokens spent per successful completion. So sure, you can blast DeepSeek for 1 hour solving a hard task, or Opus-5 for 5 minutes on the same task. Opus would ultimately be cheaper because time IS money and solving things faster is ultimately cheaper. Sure, benchmarks will show cost to be lower but people who test all of these models in real world use cases know the benchmarks are not reflective of real world use cases and Chinese models choke on problems Fable or gpt-sol will breeze through.

Re: DeepSeek V4 Flash 0731

#450

Post by Art9681 »

Cookingboy wrote: post #250
FernandoTN wrote: post #194

Earlier quoted context omitted.

I think you're overlooking the fact that for long-horizon tasks, even small errors compound over time and can lead to catastrophic outcomes. For simple queries, we have reached the threshold since the beginning of the year, and models are good enough from every provider to make a meaningful difference between one another. (ChatGPT, Claude, Gemini, Grok, MuseSpark, Kimi, DeepSeek, GLM...) The real unlock will be, and…

That doesn’t make sense. It’s not like SOTA models are error free, yet we still use them. You use Fable 5 right? If that’s good enough for you now, why wouldn’t a Chinese model that’s as good as Fable 5 but at 10% the cost be good enough in 6 months?

SOTA models are exceedingly good at self-correction in the right harness especially in domains where things can be proven mathematically. Most people are just deploying Claude Code or Codex with default settings and rub the genie lamp expecting exceptional results. Garbage in, garbage out.
Post reply on HN