frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

AI doesn't know how to forgive and cannot forget

https://tejassuds.com/blog/ai-cannot-forget
19•tejassuds•1h ago

Comments

summarybot•1h ago
It's a very good observation. I don't know if I'd call it "forgetfulness" in the context of a current [2026 edition] LLM. They are very good at remembering almost every "thing" that passes a certain token-density threshold, until they hit saturation, and then it's a sheer rockface down to the abyss of unnatural hedging and reconfirmation of basic premises. Some sort of "forgetfulness" as described would introduce more moving parts into the "inference" stage of the running of an LLM, introducing statefulness/state-tracking.
therobots927•1h ago
It’s just like me fr fr
tejassuds•1h ago
fr haha except you can forgive maybe? that's the DLC the model never shipped with
tejassuds•1h ago
unless you can't, then wait, are u agi?
therobots927•11m ago
That shi be bussin fr fr no cap
cedws•59m ago
We are a region
therobots927•12m ago
legion
throwa356262•59m ago
No, but you can reprogram it and send it back in time to undo itself.
TimTheTinker•58m ago
Ugh, another LLM-voice article.

I'm getting to the point that I almost have a physical feeling of revulsion/nausea when I read something outside an open Claude Code session that is written in an AI voice.

I'm so sick of that voice -- word choice, phrasing, style... it's hard to define and frustrating because it's both pervasive, influential on human writers including myself (I found myself using the phrase "doing a lot of work" in a Slack message about a word someone chose to use :grimace:), and presumably compounding on future LLMs.

matt123456789•57m ago
Sounds like that word was pretty load-bearing
RealityVoid•53m ago
It's probably doing a lot of heavy lifting.
Insanity•55m ago
Same here. I started describing it as the “uncanny valley” of text. It’s like a gut reaction that something is off with the text even if you can’t pinpoint it immediately.
quantummagic•50m ago
Not that it's an ideal solution, but you have all the tools you need to do something about it. You can put all your preferred writing style guidelines into an LLM and have it ingest and rewrite any text you dislike.
TimTheTinker•
josh-wrale•58m ago
I believe the title is true for humanity at large, more and more over time. These things cut both ways, for both good and ill.
tejassuds•45m ago
There is so much weight in this one comment. I feel it now!
49m ago
That's like saying a robot will be a better friend if I use a more skin-like material for its face.

The LLM-voice revulsion isn't just about particular idiosyncrasies. Another commenter put it well - the "uncanny valley of text".

alansaber•48m ago
If only there was a smoke test for this kind of thing.
myrmidon•48m ago
I concur. It's interesting that this seems so difficult to get rid of; despite actively trying to overlook the taste of LLM writing, it still grates, and this does not happen to me in the same way even with communication from half-illiterate human idiots.

It's kinda fascinating how far LLM coding/problem solving has come without making much progress on the annoyingness of their writing. Any theories?

tejassuds•46m ago
Fair hit, and I'm not going to try to argue you out of a gut reaction to the prose. If it read as slop to you, that's on me.

But, I'd rather know if the claim actually held: is there a forget decay operation anywhere in the current stack, or not? That's the part I want people to hit.

TimTheTinker•45m ago
You LLM-generated this comment too. What about you (the human, if there is one behind this account) -- what do you think?

If you don't speak English, write it in Hindi and use google translate. That will be much preferable to something written by an LLM.

Netflix in Talks to Buy Letterboxd

https://variety.com/2026/digital/news/letterboxd-sales-talks-netflix-sony-paramount-1236806379/
1•theanonymousone•45s ago•0 comments

Quills and conflict: How protection in the Strait of Hormuz is bought and sold

https://www.cnn.com/2026/07/10/business/lloyds-of-london-strait-of-hormuz-intl
1•RetroTechie•2m ago•0 comments

Could Strategy's Potential Bitcoin Sale Reshape Market Sentiment?

https://coinmarketcap.com/community/post/377668818/
1•joeymabia1•3m ago•0 comments

Show HN: WindowBunny – Window Switcher built for trackpad users

https://www.windowbunny.com
1•AbjMV•4m ago•0 comments

AI Web Design (Opus vs. Sol)

https://ploy.ai/blog/ai-web-design-with-opus-and-sol
1•brryant•4m ago•1 comments

EcmaScript 2026 (17th edition) is out now

https://blog.pstoev.com/2026/07/ecmascript-2026-17th-edition-is-out-now/
1•pstoev•5m ago•0 comments

Ask HN: Burned Tokens scented candle ideas

1•Tade0•6m ago•0 comments

Friday Night Funkin' AA Running on Windows 2000

https://www.tiktok.com/@funkingrounds_romania/video/7659481121012321558?_r=1&_t=ZN-97uypwMCEkw
1•funkinaa•7m ago•1 comments

Record Companies Push to Label AI Songs on Streaming Platforms

https://www.wsj.com/tech/ai/record-companies-push-to-label-ai-songs-on-streaming-platforms-103aa392
1•ilreb•7m ago•0 comments

JDK API Changes in Java 27

https://javaalmanac.io/jdk/27/apidiff/26/
1•theanonymousone•7m ago•0 comments

The Annotated JEPA

https://elonlit.com/scrivings/the-annotated-jepa/
2•surprisetalk•8m ago•0 comments

Show HN: Snitch – Deterministic prose claim verifier for coding agents. (OSS)

https://github.com/fristovic/snitch
2•fristovic•9m ago•0 comments

The hard part of AI root cause analysis is no longer the model

https://coroot.com/blog/hard-part-of-ai-root-cause-analysis-is-no-longer-the-model/
2•valyala•11m ago•0 comments

Memories Can't Wait–Or, How I Learned to Keep Worrying About the Web

https://zeldman.com/2026/07/06/memories-cant-wait-or-how-i-learned-to-keep-worrying-about-the-web/
2•speckx•12m ago•0 comments

Older adults know AI is slop. They just like it

https://restofworld.org/2026/china-seniors-ai-slop/
2•cdrnsf•12m ago•0 comments

Why Fear Matters in Governing Military AI

https://www.techpolicy.press/why-fear-matters-in-governing-military-ai/
2•cdrnsf•12m ago•0 comments

Man nearly sucked out of 'detached' window on Ryanair flight

https://www.rte.ie/news/2026/0710/1582735-world-ryanair/
4•theanonymousone•13m ago•1 comments

ActivityPub over ATProto

https://berjon.com/ap-at/
3•albuic•13m ago•0 comments

Offline tool to stabilize TV volume using IR control and spike detection

https://github.com/AdBusterOfficial/Adbuster--WinApp
11•Bo_Amigo_910•13m ago•0 comments

Communities will build trust for public media. Chicago is taking a leap forward

https://werd.io/communities-will-build-trust-and-loyalty-for-local-public-media-chicago-public-me...
2•benwerd•14m ago•0 comments

K2 Airways Cargo 737 Crashes into Arabian Sea

https://www.flightradar24.com/blog/flight-tracking-news/major-incident/k2-airways-cargo-737-accid...
2•teepo•15m ago•0 comments

Proton AG Services is currently experiencing some issues

https://status.proton.me/incidents/01lxtcq155lc
7•exploraz•16m ago•1 comments

Show HN: Pay-per-call MCP server with 47 AI endpoints, micropayments via x402

https://goldbean-api.xyz
2•13639366668•17m ago•0 comments

Show HN: Hacker News book recs as study guides

https://biteread.ai/collection/hacker-news-top-nonfiction
2•xxbondsxx•17m ago•0 comments

Write code like a human will maintain it

https://unstack.io/write-code-like-a-human-will-maintain-it
13•ScottWRobinson•20m ago•3 comments

Show HN: LifetimeR–A memento mori wallpaper generator (life,year,goal calendars)

https://lifetimer.vercel.app/
2•alvaro_l_torres•21m ago•0 comments

Successful Companies Go Blind

https://ianreppel.org/how-successful-companies-go-blind/
4•speckx•22m ago•0 comments

Scarf has moved away from Haskell

https://avi.press/posts/2026-07-10-after-7-years-in-production-scarf-has-reluctantly-moved-away-f...
3•aviaviavi•23m ago•0 comments

Punk, or why I don't stream anymore

https://geohot.github.io//blog/jekyll/update/2026/05/03/punk-or-why-i-dont-stream.html
8•surprisetalk•23m ago•0 comments

Go-Flavored Concurrency in C

https://antonz.org/concurrency-in-c/
3•valyala•24m ago•0 comments