frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

Open in hackernews

Ask HN: Does it still make sense to learn AI(fundamentals), if so, how?

1•shadowjones•6h ago
I've been a fullstack engineer for a few years, tinkered with AI a bit, but I've never seriously studied the fundamentals/theory of it.

From an outsider's perspective, it seems almost useless and too late to actually study AI now. (unless we're talking about how to use LLMs/agents...)

Is there any point in 2025 to go back to school and get a master/PhD or even join a bootcamp in AI/data science when by the time you'd be graduating we could have already reached the singularity?

If you wanted to work in AI and do more than just calling the latest LLMs, how would you approach it?

Comments

freigeist79•6h ago
Now is a great time to study AI. LLMs are just big neural networks , understanding how they work is most relevant. Yesterday I trained a small regression model on text embeddings for LLM preprocessing, classic ML is still very much in use. So it’s not too late at all. The fundamentals matter more than ever if you want to go beyond just calling APIs.
checker659•3h ago
https://d2l.ai

Show HN: I Turned PG's "How to Get Startup Ideas" into an Interactive Course

https://mythosgym.com/courses/startup-ideas/
1•hedayet•3m ago•0 comments

Why Juggle 3 Platforms for Your Event? Meet Demfati

1•demfati•4m ago•0 comments

I Left Quantum Computing Research [video]

https://www.youtube.com/watch?v=pDj1QhPOVBo
1•doener•8m ago•0 comments

The messy reality of SIMD (vector) functions

https://johnnysswlab.com/the-messy-reality-of-simd-vector-functions/
1•ingve•8m ago•0 comments

Tesla's Cybertruck flop is historic. The brand collapse is even worse

https://www.dailykos.com/stories/2025/7/3/2331384/-Tesla-s-Cybertruck-flop-is-historic-The-brand-collapse-is-even-worse
2•01-_-•15m ago•0 comments

AmigaLive: Front-end for FS-UAE emulator

https://www.amigalive.com/
1•doener•18m ago•0 comments

The new digital banking startup from Anduril CEO Palmer Luckey, and its value

https://www.businessinsider.com/palmer-luckeys-new-digital-bank-erebor-valuation-2025-7
1•01-_-•19m ago•0 comments

Prompting LLMs is not engineering

https://dmitriid.com/prompting-llms-is-not-engineering
20•Bluestein•19m ago•2 comments

The Tech Layoff Tracker

https://www.trueup.io/layoffs
1•skadamat•19m ago•0 comments

Unmute: Speak with a text LLM in real time

https://github.com/kyutai-labs/unmute
1•cranium•19m ago•0 comments

Built: A tool to score kids' movies scene-by-scene using subtitles and LLMs

https://tinyviewers.vercel.app/
2•kpolevoy•23m ago•1 comments

Aardvark'd: 12 Weeks With Geeks [video]

https://www.youtube.com/watch?v=YbrkZ07LKbk
1•jstanley•24m ago•0 comments

Ex-Tesla and Google Engineers Raise $4M for AI-Text Detection Startup Pangram

https://www.reuters.com/business/media-telecom/former-tesla-google-engineers-raise-4-million-ai-text-detection-startup-pangram-2025-06-24/
5•liquid99•24m ago•0 comments

Memstop: Use LD_PRELOAD to delay process execution when low on memory

https://github.com/surban/memstop
1•ingve•25m ago•3 comments

Invent provides an easy way to explore and use different AI assistants together

https://agnamihira.medium.com/why-ai-assistants-matter-more-than-ever-and-why-using-them-together-changes-the-game-d1f3ebb13752
2•agnamihira•26m ago•0 comments

The Agentic Software Engineer

https://www.dolthub.com/blog/2025-07-02-the-agentic-software-engineer/
3•ghuntley•31m ago•0 comments

300k-year-old wooden tools from southwest China

https://www.science.org/doi/10.1126/science.adr8540
1•geox•36m ago•0 comments

The Inviolable Principles of NuttX (2019)

https://nuttx.apache.org/docs/latest/introduction/inviolables.html
1•unleaded•41m ago•0 comments

Everything around LLMs is still magical and wishful thinking

https://dmitriid.com/everything-around-llms-is-still-magical-and-wishful-thinking
63•troupo•41m ago•29 comments

Show HN: Dumb STT/diction script for sway-Linux

https://github.com/winston-bosan/steno
1•csunoser•43m ago•0 comments

Iris: A neurosymbolic framework for vulnerability detection in code

https://github.com/iris-sast/iris
3•todsacerdoti•44m ago•0 comments

Using pragma Shared_Passive for data persistence in Ada

https://www.adacore.com/gems/ada-gem-20
1•ajdude•45m ago•0 comments

The Amiga 3000 Unix and Sun Microsystems: Deal or No Deal?

https://www.datagubbe.se/amix/
1•wicket•46m ago•0 comments

Being too ambitious is a clever form of self-sabotage

https://maalvika.substack.com/p/being-too-ambitious-is-a-clever-form
7•alihm•46m ago•0 comments

Cross-Compiling 10k Rust CLI Crates Statically

https://blog.pkgforge.dev/cross-compiling-10000-rust-cli-crates-statically
1•lukastyrychtr•49m ago•0 comments

Lenovo Bios Simulator Center

https://download.lenovo.com/bsco/#/
4•znpy•49m ago•0 comments

Playing Snake with water with OpenDrop [video]

https://www.youtube.com/watch?v=rf-efIZI_Dg
1•xqcgrek2•50m ago•0 comments

From Chatbots to AI Agents: Understanding Modern Agentic Architectures

https://vmayakumar.wordpress.com/2025/07/04/from-chatbots-to-ai-agents-understanding-modern-agentic-architectures/
3•mayakumar•58m ago•0 comments

Show HN: Piano Trainer – Learn piano scales, chords and more using MIDI

https://github.com/ZaneH/piano-trainer
1•FinalDestiny•59m ago•0 comments

Analysing Roman itineraries using GIS tooling

https://link.springer.com/article/10.1007/s12520-025-02175-w
1•diodorus•1h ago•0 comments