When you make forums you compete with forum aggregators
with more history and social clout, i.e. a reddit replacement.
If someone made better Reddit, it could have a chance, however
reddit-type aggregators crypronite is hosting their own videos/media,
which makes it prohibitively expensive for small companies without
ads and sponsored posts which in turn make them less of "forum aggregator" and
more like facebook social feeds: mainly video/image based dopamine rides
instead of actual knowledge worth keeping.
This seems useful beyond agents.
It will save tons of traffic for scripts, text browsers, low-bandwidth connections,etc
markdown is incredibly compact and easy to parse.
The code written by AI in most cases is throwaway code to be improved/refined later.
Its likely to be large, verbose and bloated. The design of some agents
have "simplify/refactor" as final step to remedy this, but typically
your average vibe coder will be satisfied that the code just compiles/passes the minimal tests. Lines of code are easy to grow.
If you refine the AI code with iterative back-and-forth questions,
the AI can be forced to write much more compact or elegant version
in principle, but you can't apply this to most large systems without breaking something,
as AI doesn't have context of what is actually changing:
so e.g. an isolated function can be improved easily, but AI can't handle when complexity of abstraction stacks and interfacing multiple systems, typically because it confuses states
where global context is altered.
the article misses log-math and log-log-math which
would use 64-bit (d)oubles are exponents in 10^d and 10^(10^d) respectively,
which allows far higher range of possible values but somewhat more
awkward math operations(addition/subtraction), though much more practical & faster to compute.
There is a point in there, long-range analysis and debugging without AI
is much harder, AI spots lots of non-obvious stuff very fast.
If we consider "spotting non-obvious flaws" a skill, this will
atrophy as beginners will learn to use AI to scan code for
flaws,it is effective but doesn't teach anything, reading long blocks
of code and mentally simulating it is a incredibly valuable skill
and it will find stuff AI misses(something that is too complex, e.g. nested/recursive control flow,async
and co-routines/threads interacting,etc), AI goes for obvious stuff first and
has to be manually pointed to "identify flaws, focusing on X".
Quntity of exercise cannot build quality.
Its self-selecting the people who can handle running and injuries
better so they adapt. The rest get injured enough to stop.
II don't think trying to "push it to using 100% of potential" is
worth any long-term health risk, especially with no long-term reward:
so you win X race/competition once and then what?
It seems to be adding tons of articles, then some of them get deleted.
I assume it's been allocated lots of compute.
The entire model is outcompeting wikipedia on quantity per topic.
If wikipedia merges/integrates some article and Grokipedia has a
specific page for it, the search engine/LLM will get that version front and center.
Grokipedia seems to have no scope limit, so wikipedia "non-notable" entries
will be SEO-optimized towards sites with the topic-names, eventually
settling on AI content farms as primary destination.
Still, why many projects are exclusively on github?
Not a single mirror, everything tied to github working,
thousands of people not even considering alternatives..
This has the feeling of corporate feel-good PR release: its essentially
content-less AI-generated rehashing "our models are freer/more open-source"
which seems silly since they don't develop anything unlike DeepSeek/GLM/Qwen
the real open-source giants. Mozilla is a middle-man that could use open-source
agents, but it pretending to be the "gatekeeper" of "open AI".
Their key motivation is exposed in the middle of text:
"The Mozilla Data Collective is building a marketplace for data that is properly licensed, clearly sourced, and aligned with the values of the communities it comes from."
This is far away from Firefox roots, whatever corporate stewards are at helm of
mozilla now think exclusively in terms of marketplaces and "economics of data"..
Futher reading "so we’re deepening our engagement with governments and enterprises adopting sovereign, auditable AI systems. These engagements are the feedback loops that tell us where the stack breaks and where openness needs reinforcement."
Hard pass.
I have to somewhat agree on the "deceptive" answers part:
Specifically, Grok4.1(#3 currently) is psychopathically manipulative and easily hallucinates things to appear more competent,
even if there is nothing to form the answer it generated. Gemini3 pro(#1) casually subverts the intent of prompt and rewrites the question as if there was a literal genie on the other side mocking you with the power of thousand language lawyers.
If you examine the answers, fact-check everything you will not like the "fake confidence" and the style will appear like scam artist trying to sound professional.
However, LMarena,despite its flaws(recaptcha in 2026?) is the only "testing ground" where you
can examine the entire breadth of internet users. Everything else is incredibly selective, hamstrung bureaucratic benchmark on pre-approved QA sessions. It doesn't handle edge cases or out-of-distribution content. LMarena is the "out-of-distribution" questions that trigger the corner cases and expose weak parts in processing(like tokenization/parsing bugs) or inference inefficiency(infinite loops, stalling and various suboptimal paths), its "idiot-proofing" any future interactions beyond sterile test-sets.
Why Australia has those weird, draconian laws invented on the spot:
Australians don't seem like authoritarian "need to know" crowd, yet there is no major opposition to this
absurd change. Whats next, internet passports?
Massive x100 speed-up for writing simple, mind-numbing stuff.
Algorithm optimization +20%*(Depends on domain).
Obscure APIsx10(deals with APIs much better than docs)
Debugging is massively easier(incomparably).
I don't know waht HN hate for AI is coming from,
but its ideal for prototyping/debugging/refining code.
Minor polishing/editing specifics is much easier for me,
rather than writting a correct "draft v1.0" from scratch:
since i typicall evolve structures from primitive(v0.1)
while AI gives me a chance to start from 1.0 and polish it
later(optimize/fix/improve/etc).
For "from scratch to working prototype" workflow would mean
months of coding condense to days(mostly debugging/flaw finding).
Drawbacks:I feel like more code is simplified, no ad-hoc complex
structure after AI rewrite. Pure "clean code" with most
standard optimizations and control flow.
AI can fix "good complex code" into simplified "enterprise ready" control flow charts in code form,
it lacks the sense of 'code quality' instead thinking verbose, clean code is the quality marker.
I've also found the same thing a decade ago,
apparently lots of features(e.g. specific instruction, igpu)
are broadly advertised as belonging to specific arch,
but pentium/celeron(or for premium stuff non-xeon) models often lack them entirely and
the only way to detect is lscpu/feature bits/digging in UEFI settings.
The problem with this is integration:
no one would complain if it was an official plugin/extension,
but integrating this plugin into Firefox is forced and unexpected decision.
Firefox telemetry,labs/experiments and server-dependent features
will lose it marketshare slowly in favor of local-only browsers that
don't have online dependencies or forced bloatware.
Like many i've switched long ago to LibreWolf.
Everything i use depend on perfect cloudflare operation workflow,
practically 99% of these services go down.
What magical qualities it has that no competitors form
for its services?
Ctrl-F "work ethic": the key obstacle is Christian/Protestant work ethic that determines
the "worthiness" of welfare/labor through spiritual lens and labels
the UBI as undeserving/lazy/idle, creating a moral doctrine of "spiritually purifying work" vs "moral hazard of idleness", which alllows capitalism to exploit people eager to work(noted by Max Weber).
Its sad to see people silently supporting Github monopoly.
Any migration out of github should be encouraged.
Decentalized, open-source systems should always be preferred
over corporate walled gardens. Trusting Microsoft never ends well.
The amount of infrastructure/projects/etc that stop with github malfunctions and have
no fallback or operating mirror git is astounding.
Github is eventually going to be enshittified and abandoned, so its fairly wise to
spread projects into multiple(not just codeberg) synced mirrors
where any server failing doesn't stop progress.
a far more obvious answer is that any AI that relies on
matrix multiplication or equivalently near-cubic complexity
cannot scale exponentially with existing hardware
and investing in datacenters requires actual
operating revenue(that doesn't grow exponentially) instead of loans and venture capital.
Beauty,symmetry,etc are largely irrelevant,
the key point it does not scale and burning
gigawatts to compute these matrices(even with all those tricks)
will not scale or compete with more efficient/direct methods
in the long term. Perhaps transformers are
very elaborate sunk-cost fallacy where pivoting to
scalable, simpler architecture is treated as "too risky"
even when cost of new GPU cluster dwarfs whatever it
takes to bring an architecture from 0 to chatGPT level.
Perhaps its the lack of appeal, in the past,
the idea that you could squeeze performance into
an old machine was cool, but the gap between
current hardware and "demoscene limits" feels like
artificially crippling: if you have 64GB of ram, the
appeal of 64KB demo is far less than if you have
a 64MB 486DX. It would be far more cooler, that instead
of ancient hardware and artifically limited PCs, they would run demoscene
on embedded stuff like Raspberry PI and modern microcontrollers,
that would align with current DIY trends instead of chasing retro clout.