News for hackers. Stuff that matters.
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by cmiles8 on Wednesday August 05, 2026 @09:15PM (#49191212)
The big lab models are academically interesting but business wise they seem toast long term. There’s no way for these model companies to compete when the models are becoming a pure commodity and others offering options that are orders of magnitude cheaper.

It’s not that the big labs couldn’t theoretically just also put out 100x cheaper options but their business model requires them to generate huge revenues from higher priced tokens or they’ll implode.

Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by typ on Wednesday August 05, 2026 @11:09PM (#49191934)
Electricity becoming a commodity doesn't necessarily make nuclear reactors become commodities. And as we can see, the publication of open-weight models doesn't really diffuse the know-how of training SoAT models as we expected. Even the major cloud providers are still unable to build competitive ones from scratch on their own. That's very different than the traditional FOSS ecosystem, where everyone can copy, learn, and evolve on one's own (like, Bitcoin -> altcoins).

On token pricing, I think it's very much bottlenecked by hardware (the aggregate of compute) rather than the number of competing models. Assuming that the ceiling of the token price is determined by the economic value a unit of compute can provide, then the less efficient ones would be priced out of the compute allocation. It's not consumers bidding up a limited number of different models, but more like tokens of different models bidding up the limited computing resource. Less-intelligent tokens (which are generated by weaker models) are crowded out by smarter tokens from the limited compute. My prediction is that we'll see a meaningful downward pressure on token prices only when the new batches of next-generation hardware get mass-deployed.

Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by podocarp on Thursday August 06, 2026 @05:34AM (#49194491)
I don't think the big companies not making Sota models being an indicator that open source failed. If all semiconductor patents expired tomorrow it would still be prohibitively difficult to manufacture your own CPU at home or at work.
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by typ on Thursday August 06, 2026 @08:16AM (#49195621)
Me neither. But open weights are not open source in the first place. It wouldn't be surprising that the raw intelligence of an LLM model has little to do with the architecture (which is the part most published). The secret sauce, like how/what data gets curated and cleaned, has rarely been talked about. So, even if the labs upload their weights, competitors are very difficult, if not impossible, to replicate the technology independently.
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by childintime on Thursday August 06, 2026 @10:23AM (#49197094)
> Electricity becoming a commodity doesn't necessarily make nuclear reactors become commodities.

Bad comparison? Nuclear reactors become obsolete. The only reason to build a nuclear reactor is as a government vanity project, just to get the bribes you want, or weapons grade material. Dirty tech.

Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by LunaSea on Thursday August 06, 2026 @10:34AM (#49197246)
And you know the small benefit of having stable, large scale and low CO2 emission electricity production.
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by childintime on Monday August 10, 2026 @12:47PM (#49246241)
You have put on rosy glasses. Solar begins saving CO2 immediately, while nuclear reactors begin emitting CO2 before they are ready. After a 10-15 year investment in concrete that results in around 4x the CO2 emissions.. they are obsolete by the time they start producing. And additionally they produce still nuclear waste, nobody wants to have them nearby, and they are a strategically high risk in multiple ways (military, financial, environmental). They are vanity projects.
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by LunaSea on Saturday August 22, 2026 @10:25AM (#49400082)
Vanity projects that make France produce electricity with 25x less CO2 / kWh than Germany.
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by kumama on Wednesday August 05, 2026 @09:32PM (#49191321)
castform founder here. despite our bet on fine-tuned smaller open-source models, i'm still quite bullish on the big labs. i think scaled closed models will continue to dominate for more general purpose use-case like codegen, search, etc. but intelligence has lots of long-tail applications and i think for these longer-tail applications, finetuned custom models will rule
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by sroussey on Wednesday August 05, 2026 @09:49PM (#49191450)
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by kumama on Wednesday August 05, 2026 @09:53PM (#49191479)
this is really cool. i'm sure some of the larger e-commerce companies are already using clickstream data to tune better query rewriters/ranking models
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by podocarp on Thursday August 06, 2026 @05:39AM (#49194532)
Sorry but what is neurosymbolic model, it's not explained I think
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by sroussey on Thursday August 06, 2026 @03:10PM (#49200970)
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by fennecfoxy on Friday August 07, 2026 @10:31AM (#49211039)
I would also like someone to clarify this for me. I (think) I understand what people mean by a neurosymbolic model however common definitions are a bit funky.

If it's neuro (llm/transformer similar) symbolic (symbolic with hand crafted rules) then, imo, that's no different from current tool calling harness implementations and I'd love for someone to explain it further to me if I've misunderstood.

If it's neuro (llm/transformer similar) symbolic (symbolic rules that have also been learned via training) then yeah I can understand how it's a distinct concept.

But every definition I've seen tells me that a standard harness of:

query->llm->[tool call symbol + tool name + tool params] aka symbols governed by logic, including param/arg validation->llm->etc

Already meets the requirements for being "neurosymbolic"...

Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by nosduhz on Wednesday August 12, 2026 @02:21AM (#49268492)
Can totally see your point (and there are probably cases) where people are applying a harness on top of an LLM and calling it neurosymbolic.

In our case with Ontology, it's the latter! Symbolic rules learned via training + constantly updating autonomously.

Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by nosduhz on Wednesday August 12, 2026 @02:16AM (#49268454)
Woah thank you for posting!
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by esafak on Wednesday August 05, 2026 @10:30PM (#49191725)
Small models aren't going to take people's jobs. Agents using large models might. That's how their numbers make sense.
Re:Beating GPT-5.6 Sol on retrieval with 100x cheaper open m... by cmiles8 on Wednesday August 05, 2026 @11:22PM (#49192038)
Except there’s been very little evidence of this happening or about to happen, which is turning into a huge problem for the big labs.