segmondy

8 ms·

151.
by segmondy 4mo ago
Fable to Opus is the biggest scam ever, I suspect a class action lawsuit will be landing soon.
152.
by segmondy 4mo ago
Sep 24th 2026. 10:07am EST
153.
by segmondy 4mo ago
Buy a $50 used chromebook and take it in.
154.
by segmondy 4mo ago
A bit too late for that, most of them have already dumped most of their codebase and IP into cloud models.
155.
by segmondy 4mo ago
The rumors I heard was that once llama3 became successful, everyone that had influence wanted to attach themselves to it and they did, destroying the original team and the culture in the process, by the time llama4 landed the smart ones wer
156.
by segmondy 4mo ago
The failure that is llama4 needs to be studied. Meta was kicking ass with llama3.x and then something happened, something really went wrong. what happened between that time and llama4? I think it happened after llama3.1, llama3.2 was n
157.
by segmondy 4mo ago
Culture
158.
by segmondy 4mo ago
Why should salaries catch up to heady levels of SV? There are very few places in US that can match SV in salary.
159.
by segmondy 4mo ago
Fine by me, as a 25 years user. More is not necessarily better, see redis.
160.
by segmondy 4mo ago
no matter what, success or not the bubble is going to bust, it has inflated so much, the risk of it deflating slowly is a pipe dream. if AI doesn't turn into AGI, global financial crash. if AI turns into AGI and tons of people are o
161.
by segmondy 4mo ago
these are being mapped, i will like to see a GPS route option to avoid flock cameras, but I suspect they will end up flooding all intersections.
162.
by segmondy 4mo ago
my point is that if someone is releasing these without access to mythos/gpt5.6, then it can spook to govt and cause them to keep suppressing access to those models. if i'm competing with anthropic or openai and feel they have pul
163.
by segmondy 4mo ago
With the crazy hardware prices. It's time to go lean. Embrace lean software and learn to do with less or pay the price with your hard earned $$$.
164.
by segmondy 4mo ago
As we can see again, this has nothing to do with distillation, yet for every gain Chinese labs make, the US labs will accuse them of theft. Yet they are constantly innovating.
165.
by segmondy 4mo ago
What if this person is from an AI lab that really wants the govt to keep suppressing Mythos/Fable & GPT5.6? It's what I would do, the timing couldn't be any better.
166.
by segmondy 4mo ago
Well, I can't be mad if I ever get accused if Plank has no chance.
167.
by segmondy 4mo ago
late 2027 is a very long time.
168.
by segmondy 4mo ago
What we want to really know is the cost of Ultra Mac Studio 512gb if that will happen or 256gb.
169.
by segmondy 4mo ago
The bottom line has not shown that to be the case.
170.
by segmondy 4mo ago
For all the complaints about Anthropic many of you still give them your money! Stop using it. I don't care if they claim they are the best model. I stopped paying OpenAI and Anthropic 2 years ago once they started going for regulato
171.
by segmondy 4mo ago
Serious question, do you realize that the NSA are mere mortals? Do you realize how much it takes to train a model? Does the NSA make their own chips or planes? The NSA buys a lot of technology because they can't make their own.
172.
by segmondy 4mo ago
The really crisis is that we are now in a phase of ("AGI" or Bust). That's it. Unless we get a self improving autonomous AI heading towards "singularity" then it's all going to blow up. The amount of money
173.
by segmondy 4mo ago
AI vibes all the place!
174.
by segmondy 4mo ago
Here's a thought experiment for you. Let's say you can run 1000 agents at 10,000 tokens a second. Do you think you are going to be more productive than someone running at 6tk/sec with the same model? Incase it's not c
175.
by segmondy 4mo ago
No one locally runs full load all day. The only way to see that is if you're training. We are talking about inference. I limit my GPU to 300watts. You can limit them down to 200w. Since everything is not in GPU and the bottlene
176.
by segmondy 4mo ago
512gb 2400mhz ddr4 ram = $1600 not $5000. https://www.ebay.com/itm/188284985172 You can get creative and source 2-3 2080ti 22gb from China for about $250 a piece. You can either be resourceful and find a way or find
177.
by segmondy 4mo ago
I have been putting up with it forever. We are spoiled by MixtureOfExperts. Folks were delighted to run llama3-70B at such speed. We were happy with 15-20tk/sec with 8b models, and if you could run llama3-405B at 1tk/sec you
178.
by segmondy 4mo ago
checkout llama.cpp, the entire point of the project is for us mere mortals and GPU poor.
179.
by segmondy 4mo ago
you can use system ram with a system like llama.cpp which offloads to system ram. token generation is a function of system bandwidth, the faster the bandwidth the better. so I'm on 8 channel 2400mhz. if I had a 12 ddr channel, I wou
180.
by segmondy 4mo ago
You're talking about agentic workflow. Agentic is cruise controls. Race car drivers shift manually for more precision and to go faster. If the only way you know how to code with AI is agentic, then you are putting yourself on a crut

More ›