spprashant

Everything since that launch to this release has been a PR disaster for Anthropic.

They are trying so hard to make AI do human jobs instead of focussing on opportunities where AI is special suited. Do you really want your super intelligent token muncher to be clicking browser tabs all day?

I can't believe they are yanking tool access instead of just reducing the token quota or simply pulling Opus 4.7 access. To be fair even that would be poorly received, but at least people would have a choice of working within limits. Claude Code is their real winner, and a great ramp for newcomers coming into AI assisted development. They are playing straight into OpenAIs hands.

None of those seem like they had an capital investment equivalent to 1% of the GDP. Apparently railroad was the only technological investment that was higher than AI when measured as a percentage of the GDP.

Yes. But unfortunately that domain suffers from ambiguity which LLMs are bad at.

Medical treatment has never been about asking questions and getting perfect answers. Excellent doctors and nurse practitioners have a great intuition for which questions to ask based on cues during patient assessment.

I think all misgivings about AI would go away fast, if it solved one important problem for humanity. Carbon nanotubes for space elevators, sustainable nuclear fusion, or something in that ilk.

The positive effects were immediate, and measurable. The negative effects are delayed, and hard to quantify without all the advancement in climate research since then. If everyone in 1920 knew a 100 years from now there would be climate crisis to reckon with, perhaps a few things would have changed along the way.

Today we have a much better understanding of the world, so we have the means to think down the line of what the negative effects of LLMs and course correct if needed.

Can you give me a rough example of what you prompt? Just asking for info. I use Sonnet 4.6 and have a hard time hitting capacity.

I don't know if Gastown will work out. But it is quite a bold take. I am interested to see how it plays out. I suspect they will eventually roll back some of the stringent "no code reading" approach in favor of observability as the community grows.

I think it behooves us to be selective right now. Frontier labs maybe great at developing models, but we shouldn't assume they know what they are doing from a product perspective. The current phase is throwing several ideas on the wall and see what sticks (see Sora). They don't know how these things will play out long term. There is no reason to believe Co-work/Routines/Skills will survive 5 years from now. So it might just be better to not invest too much in ecosystem upfront.

I don't think the disconnect is very surprising to the "insiders".

Your Dario's and Sam's know exactly what they are doing. They know it's going to cause a lot of job displacement, even if the technology isn't perfect. They are trying to get the C-suite elite hyped up about it, and the hyperscalers are along for the ride as well. There's so much money to be made.

They could not care less about what joe schmoe on the street thinks about it.

First I am hearing of Gateway Arch national park, and I am very confused why it's a national park?

At this point, I almost feel bad that people are piling on Garry Tan. Almost.

And how does BTC play in this scenario?

I tend to prefer the ARC-AGI benchmarks for the most part. But it's always interesting when a new version drops, all the frontier models drop less than 20% or something. And then in the next few releases they get all they way up to 80%+. If you use the models it doesn't feel like those models are that much more generally intelligent.

Most frontier models are terrible at AGI-3 right now.

These models are already great no question, but are they really going be that much more intelligent when we hit 80% again?

I have a Kindle from 2014 still going strong. I guess it ll be bricked in a couple of years.

The problem is once you accept that it is needed, you can no longer push AI as general intelligence that has superior understanding of the language we speak.

A structured LLM query is a programming language and then you have to accept you need software engineers for sufficiently complex structured queries. This goes against everything the technocrats have been saying.

Sounds like a good effort. They are choosing to focus on multi-modality - perhaps they are taking a different route here to Anthropic.

I don't like that I need to login to my FB/Instagram account to access this.

In Multimodal yes, but Opus is definitely edging out in Text/Reasoning and Agentic benchmarks.

I think the general skepticism is because they are late to race, and they are releasing a Opus-4.6-equivalent model now, when Anthropic is teasing Mythos.

I wonder if this is why the tech cartel is buying up all the hardware?

If the average user gets convinced they could run LLMs for cheap at home, you cannot trap users in your walled garden anymore.

We final have the answer to the question, when do these labs stop giving away intelligence to the general public for $20 a month?

Selling shovels in now worth less than taking all the gold for themselves.

There's a little explained if you hit the (?) at the bottom. They are taking the monthly inflation value and calculate it per tick.

You know what, it's not as bad as I was thinking.

Well the important thing is they have a lot more data of people actually using their models. They have read billions more lines of private repos and implemented millions of patches, all of which is feeding into the newer models.

More importantly it understand what behaviour people tend to appreciate and what changes are more likely to get approved. This real world usage data is invaluable.

I am not smart with stock legal-ese but I pasting something I found in a different article here.

> To balance index integrity and investability, Nasdaq proposes a new approach for including and weighting low-float securities (those below 20% free float). Each low-float security’s weight will be adjusted to five times its free float percentage, capped at 100%. Securities with more than 20% free float will continue to be weighted at full, eligible listed market capitalization, while those below 20% free float will be weighted proportionally to preserve investability.

> The rule reportedly includes a 5x float multiplier for low-float stocks, which would require passive vehicles to treat SpaceX as if it had significantly more tradable shares than actually exist, essentially forcing funds to chase the price.

It sounds to me like a way to increase demand for low float stocks by treating the float higher than it actually is. Glad to hear the explanations about this.

You are not supposed worry about the mapping. You trust the website to help decode it. You just remember the sentence. It's a little like what3words for coordinates.

The rationale being you are more likely to remember grammatical cogent sentence, than a random string of alphanumeric characters. Although I will agree that the generated sentences don't seem easy to remember. So I doubt it's utility.

Its simple, but its not easy is what I would say. Once you figure out the meta, you can work out most of it.

I played the demo, but it definitely took me a minute to grok the rules.

I don't know if this is how we want to measure AGI.

In general I believe the we should probably stop this pursuit for human equivalent intelligence that encourages people to think of these models as human replacements. LLMs are clearly good at a lot of things, lets focus on how we can augment and empower the existing workforce.

Feel bad for the ATC officer. I hope they can find it in them to forgive themselves.

I only just started using it at work in the last month.

I am a data engineer maintaining a big data Spark cluster as well as a dozen Postgres instances - all self hosted.

I must confess it has made me extremely productive if we measure in terms of writing code. I don't even do a lot of special AGENTS.md/CLAUDE.md shenanigans, I just prompt CC, work on a plan, and then manually review the changes as it implements it.

Needless to say this process only works well because: A) I understand my code base. B) I have a mental structure of how I want to implement it.

Hence it is easy to keep the model and me in sync about what's happening.

For other aspects of my job I occasionally run questions by GPT/Gemini as a brainstorming partner, but it seems a lot less reliable. I only use it as a sounding board. I does not seem to make me any more effective at my job than simply reading documents or browsing github issues/stack overflow myself.