robswc

I guess I'm missing something but I don't see how this isn't just:

1. Creating more work 2. Shifting the existing work to specification

There's not really such a thing as a perfect specification. One of the things AI is taking away is character by character feedback cycles. Can't count how many times I'll be writing software and think in the middle of writing a return, realize this is probably not what I should be doing. I'm not sure how you would really replicate that experience when creating specs... nobody is going to sit and meditate on their specs for hours before writing them.

I have been ringing this bell for years.

When I first started software, I thought it was the weirdest thing in the world for everyone to go "ok next step, download 1000 different pieces of code from 100 different places."

The only thing missing was the financial incentive, and ease of entry... and man if crypto + AI haven't got that combo down.

Sure. I can see how the problems being solved would be very different, and would require "re-training." I'm wondering though how much is it the LLMs and how much is it the humans driving them? Could this have been done with the models currently accessible? Could I have just prompted "solve this conjecture" and hand the work over to a PhD to verify? Or is it "unlimited budget" + "super smart ppl prompting" + "SOTA model" ?

I guess I just see this like how NASA went to the moon. In theory there wasn't a lot stopping us from going besides a good reason (and money). OpenAI has a good reason (PR) and money, so they're doing it. Will these things stop coming out after funding and motivation is no longer there? Or by next year will we have access to this model and anyone can solve any of these open conjectures with time + money?

> If you have suggestions on how to name our products better, I'm all ears. Our intention is to make things clear and simple, even if we miss the mark sometimes.

I'm not the one you were talking with but catching you is godsend since I just ran into a lot of confusion on the naming myself. I think some of it has to do with how they are labeled in the app.

I think simple numbers in naming is great, even 6.1 is fine bc I understand that if I want the latest model, I just need to look for the biggest number. Where I got confused is in the "work" tab, I saw 'Astra' is set as the default. I see the label for it is:

"Frontier intelligence for your most demanding tasks."

To me that reads "this is the 'best' model." Then I see Sol-6.1 which is labeled:

"For coding and every day work."

Then Terra-6.1 is:

"Terra 6.1 Codex Model" (tbh this one might be the most confusing, I'm aware of codex the product but Sol is the one labeled good at coding)

From my POV as a basic plan-based user wanting to do something like...

> design a robust event-driven system, here's the details [...]

I honestly have no idea which one I should use.

Would Astra-6 produce a better design than Sol-6.1? Is the '.1' a big step up from Sol-6? (I'm guessing no but I truly have no idea). I would assume the answer is Astra-6 would produce the best one, even though the labels don't imply it is necessarily geared towards coding?

But tbh the "default" seems to always work for me... and after reading your comment, typing this one out, being forced to think about it... it doesn't seem too bad given your challenges in communicating the purposes of each model and wide audience of consumers. I'm also sure anyone could find a million problems with decisions I've made in software I've built so I try to not be too harsh :)

I would trade (as if I have any say lol) every single breakthrough they have made in math/CS for _any_ good progress towards a universal cancer cure (very theoretical, I know).

I'm kinda hoping they are investing just as much into that type of research. Its their money but man... I'd say 999/1000 people are unaffected by all these breakthroughs. They probably have no idea tons of legendary math/CS conjectures were proven true/false. If a breakthrough in cancer research were to be made, it would very quickly start effecting everyone on earth.

Exactly my thoughts.

I find that unless the outcome I'm looking for from the model is almost binary and easily verifiable, it is basically faster to do the work myself.

Assumptions made by the models multiply and by the non-deterministic nature (dbf, similar to work by humans) I'll have something end up light years away from where it should be. During a normal workflow, I have to contend with every line I write forcing a very tight feedback loop... I suppose I could also sit there and spend 9 hours writing out a "perfect spec" but we've already tried that with waterfall development!

So, not sure if its the models, rather the class of problems I'm applying them to... but it is a bit ironic they are allegedly able to solve these conjectures but not stop to ask about assumptions.

I'm incredibly skeptical about everything these frontier labs put out. Results are results though.

However, could someone with a firmer grasp on the problems weigh in on how much are these recent results are function of cutting a virtually unlimited budget for some very smart people? I believe the LLMs certainly help iterate on ideas faster... but I am aware of very impressive work that goes on in quantitative finance. Stuff that would not really be published. That stuff is the result of paying smart people millions to produce results which produce $$$.

I have to wonder, would this have been solved if we simply gathered tons of smart people and paid them a ton of money to solve it? Or is it exclusively because they had access to LLMs?

Could they have achieved this result with previous versions of ChatGPT?

> So vote for a government that will do it.

There will almost never be a government that does this because the incentives are not there.

There are more incentives in prolonging a problem than there are in solving it.

> Or run for office yourself

Winning an election against career politicians is difficult and next to impossible. There is virtually no way to do this without having backing. As soon as you have backing you have new incentives. The purpose of a career politician and the machine behind them is to win elections. That may or may not involve solving problems that matter to constituents.

> I thought "the sentiment" was "tax the rich because they can afford it".

Not "because they can afford it" but "because it doesn't hurt." The "hurt" I'm referring to isn't the wallets of the rich, rather the existing revenue, the balance of things, society, etc.

The true sentiments will always depend on who/what is being taxed and who/what is advocating for the tax. Almost everyone I see supporting this particular tax is supporting it more as a means of "punishment" that can be levied against the rich.

> Taxing people with money makes the most sense.

Sure. Not sure how you would tax people without money... but it is always going to come down to "how much" and "what for", more than anything. Then you have to contend with downstream effects. People with money can and will move if they feel they are being taxed too much. This is no different than setting the price of a widget. You can always raise it but then you will have less buyers.

Probably. However, I've mostly been trying to watch older movies and I think paramount has some sort of monopoly on them.

I haven't seen the episode but I can imagine the critique is pretty valid. However, sometimes it goes too far in the opposite direction. Sometimes things _are_ actually worse now :)

There's a lot of nostalgic things that do in fact suck, I have no problem admitting that. Subjectively, I will die on the hill that movies overall are getting worse though. They are all starting to feel over-produced and sanitized. Objectively, the time between seasons of hit TV shows is increasing, we can measure that as a fact. Not rose-tinted glasses.

Honestly I think it's because everything sucks these days. I know there is a lot of survivorship bias going on but _something_ has changed. I have recently been on a 1980-2010's movie binge. It seemed like movies before were raw and real. Now everything feels like it's plastic and made in a lab, dumbed down to the lowest common denominator.

Don't even get me started on shows. There are a ton of great ones but now they have 2-4, sometimes even 5 year gaps! The "increased" quality doesn't warrant it! I know people say "well now they're movie-grade" but I really could not care less. I don't need every episode of every show to be blockbuster quality. Just make it not suck! I would love to go to the movies and buy all the subscription services but it's simply not worth it! /rant

> Paramount Plus has nothing else.

This is funny, because every damn time I find something I want to watch its on paramount+.

There has to be a paradox/law named after this at this point. "The thing you're trying to watch will always be on the service you don't have."

> That's the sentiment behind every tax

Says who? Taxes don't exist to "punish" the rich or successful businesses. They exist to pay for things. If someone is mostly looking forward to the taking of money rather than the building of things with that money, they are being ridiculous.

In general, the foundations upon which it was built.

In specific, off the top of my head:

- you can't group repos

- CI/CD is far behind what GitLab offers

- Review apps (though I think now you can do this on GH with other tools)

- Can't mark PRs as drafts by default in GH

- better metric tracking, easier to see a "snapshot" of where something is at on GitLab

Tbf, GH may have closed some gaps over the years but when I was bouncing between the two GH always felt a lot more rigid and further behind than GL. It did have fantastic support for open source though, and has a few unique and fun things like the `io` site stuff. Overall though if I could choose which to work in for serious projects I'll always choose GL.

GitHub is such a mess. They seem to be doing everything and nothing. GitLab is 100x better, in terms of DX. While I still have a soft spot for GH, its frustrating to use it when GL exists.

I can agree. I believe the R&D component is very real but will take/has taken a backseat to the business.

That would be great. Unfortunately as soon as governments get more money they will spend it which makes it increasingly harder to spend less.

Sure, not claiming it is the same category. Just the sentiment of "it's just taking a little from the rich so it won't hurt" is similar. The main emotional motivator behind both is to take from the rich.

I honestly don't see how this outcome isn't all but guaranteed. It is actually almost comically self-destructive. Taxes based on the same idea of (just take a bit from the rich) destroyed the Swedish financial markets (basically forever).

https://en.wikipedia.org/wiki/Swedish_financial_transaction_...

> As a result, revenues from these taxes were disappointing. For example, revenues from the tax on fixed-income securities were initially expected to amount to 1,500 million Swedish kronor per year. They did not amount to more than 80 million Swedish kronor in any year and the average was closer to 50 million.[4] In addition, as taxable trading volumes fell, so did revenues from capital gains taxes, entirely offsetting revenues from the equity transactions tax that had grown to 4,000 million Swedish kronor by 1988

So off by an order of magnitude in their estimates _and_ nuked revenue from the capital gains taxes. Overall probably one of the dumbest ideas ever executed because it was so easy to forecast the second order effects.

It depends on the day and how honest people feel. If it were truly about numbers they would probably find more success in getting their government to spend smarter.

I am just amazed at how many people are blind to this.

Every single time one of these labs does something it is almost always sensationalized. They absolutely cannot refrain from describing how incredible their models are, it has to be injected into everything. I get this is basic grow-your-company stuff but it is getting obnoxious.

Sorry for the late reply.

If I look in the "Library" at images I have generated, scroll down a bit and click the image, all the images turn black/empty, that image then shifts to the top of the page.

Putting it at the top is fine, but it should be done _after_ one closes out of the library.

> Didn't this exact scenario just happen like 3 months ago with Instagram?

Hey now, they probably paid millions of tokens to ship that feature.

Not sure I can agree, but the idea is interesting.

While the set is finite, software comes and goes. Just look at the tools to build software over the years.

> Many software firms have been allowed to bloat up do to the nature of the business.

I can agree with this. It will be harder to justify engineers sitting on laurels. Google could afford it before but I can see how it will be under more scrutiny now.

I can't even go a day without running into super frustrating bugs on the ChatGPT app.

Totally agree. IMO it's mostly behind the scenes stuff though.

I think some of the most impressive stuff to show someone from 2016 might be LLMs doing a summary of complex topics, followed by self-driving (actually) cars and intricate robots... but other than that there is no "future magic" touching the lives of every day people going on.

Actually I'm pretty sure if you showed someone the latest and greatest from these big AI companies, these AI companions that can book flights, people would go "Oh, so its like a smarter Siri" and proceed to not care at all lol.

Fair point. I have to wonder how much is because of AI and how much is from ZIRP stuff though.

This is sort of my take. You could look around in 2016 and 2026 and honestly not much has changed... at least at first glance. There is a missing piece of the puzzle.

I'm noticing this even in industry. Yea, we can build and iterate on ideas 100x faster but has this turned into anything meaningful? Not really... Hell, look at the big companies spending billions on inference... nothing shipped and things are still buggy as ever. Where is everything?