HackerLangs
TopNewTrendsCommentsPastAskShowJobs

pulkas

no profile record

Submissions

[untitled]

1 points·by pulkas·há 20 dias·0 comments

The Last Invention of Man

nautil.us
5 points·by pulkas·há 22 dias·2 comments

[untitled]

1 points·by pulkas·há 25 dias·0 comments

Ask HN: How can I use other LLMs as tools, not slaves but collaborators?

2 points·by pulkas·há 9 meses·2 comments

Synthetic: All-Inclusive AI Platform with 19 Always-On Models for $20-60/Month

synthetic.new
1 points·by pulkas·há 10 meses·1 comments

comments

pulkas
·há 17 dias·discuss
I think the next movement is heading to multi model orchestration.

https://developer.nvidia.com/blog/train-small-orchestration-...
pulkas
·há 18 dias·discuss
[dead]
pulkas
·há 20 dias·discuss
when will WASM integration become widely accepted? we need realtime canvas draw like oldschool desktop ui/ux.
pulkas
·há 22 dias·discuss
Tegmark's 2017 piece on AGI and existential risk — still very relevant given recent developments in AI.
pulkas
·há 25 dias·discuss
waiting for the anouncement: cursor for grok heavy users.
pulkas
·há 28 dias·discuss
[flagged]
pulkas
·mês passado·discuss
article is too old. who is using gpt-4o today?
pulkas
·há 2 meses·discuss
they are testing the limits of ruby on rails.

"Ultimately, if more companies treated the framework as an extension of the application, it would result in higher resilience and stability. Investment in Rails ensures your foundation will not crumble under the weight of your application. Treating it as an unimportant part of your application is a mistake and many, many leaders make this mistake." https://github.blog/engineering/architecture-optimization/bu...
pulkas
·há 4 meses·discuss
what happens if bitflip occurs while you are detecting bitflip?

bitflippin...
pulkas
·há 4 meses·discuss
your better make it as chrome extension.

And i dont know why you wanted me to give you all my permissions?
pulkas
·há 5 meses·discuss
There’s something almost nostalgic about it: cutting-edge technology wrapped in layers of bureaucracy that feel lovingly inherited from another era.
pulkas
·há 5 meses·discuss
This violates the Claude Code subscription terms of service, so please be careful.

This project violates Claude Code's Terms of Service by automating Claude to create an unattended chatbot service that responds to third-party messaging platforms (WhatsApp, and what you add ...).

  The exact issues:

  1. Automated, unattended usage - The system runs as a background service (launchd) that automatically responds to WhatsApp
  messages without human intervention (src/index.ts:549-574)

  2. Building a bot service - This creates a persistent bot that monitors messages and responds automatically, which violates restrictions on building derivative services on top of Claude

  3. Third-party platform integration - Connecting Claude to WhatsApp (or other messaging platforms) to create an automated
  assistant service isn't an authorized use case.

  The README itself reveals awareness of this issue at line 41:

  **No ToS gray areas.** Because it uses Claude Agent SDK natively with no hacks or workarounds, using your subscription with your auth token is completely legitimate (I think). No risk of being shut down for terms of service violations
  (I am not a lawyer).

  The defensive tone ("I think", "I am not a lawyer") indicates uncertainty about legitimacy. While using your own credentials doesn't automatically make automated bot services compliant—Anthropic's TOS restricts using their products to build automated chatbot services, regardless of authentication method.

  The core violation: transforming Claude Code into an automated bot service that operates without human intervention, which is explicitly prohibited.
pulkas
·há 5 meses·discuss
The Mythical Man-Month, revisited
pulkas
·há 6 meses·discuss
@kagi scrape here.

https://kagi.com/smallweb
pulkas
·há 6 meses·discuss
zigdom all-diy
pulkas
·há 8 meses·discuss
darklang team loved ocaml too. in the end they ended up using f#.
pulkas
·há 10 meses·discuss
Synthetic offers a subscription-based alternative to pay-per-token AI pricing. Instead of tracking usage costs, you get unlimited access to all their "always-on" models for a flat monthly fee.

What's included:

- 19 always-on models with both UI and API access

- LoRA fine-tuning support (FP8 precision, up to rank-64)

- Embedding models at no extra cost

- Standard: $20/month (135 msgs/5hrs), Pro: $60/month (1,350 msgs/5hrs)

Technical specs:

- Always-on models: No quantization (full precision)

- On-demand models: BF16 precision (FP8 for Jamba-based models only)

- LoRAs: FP8 precision, rank-8 to rank-64 support

- On-demand GPU pricing: 80GB at 3¢/min, 48GB at 1.5¢/min, 24GB at 1.2¢/min

- On-demand context limit: 32k tokens

Complete always-on model list:

DeepSeek: R1, R1-0528, V3, V3-0324, V3.1 (all 128k)

Meta Llama: 3.1-405B/70B/8B, 3.3-70B (128k), 4-Maverick-17B (524k), 4-Scout-17B (328k)

Others: Kimi-K2 (128k/256k), GPT-OSS-120B (128k), Qwen2.5-Coder-32B (32k), Qwen3-235B variants (256k), Qwen3-Coder-480B (256k), GLM-4.5 (128k)

Additional features:

- LoRA support for Llama 3.1/3.2 base models

- Embedding model: nomic-ai/nomic-embed-text-v1.5

- Any HuggingFace model available on-demand

Links: https://synthetic.new/ | With referral: https://synthetic.new/?referral=9oxapskWLeOrDT5