He's clearly focusing on being on the news to provoke emotions on people, preparing for the Anthropic big blockbuster IPO.

Dario, Sam Altman and others should instead be praying every night that this will take us to the Singularity VERY SOON, because if it doesn't, the entire country will blame them for absolutely destroying their retirement and the US' economy once this AI Datacenter bubble pops.

My take: no

They are still pretty much competing for software development, you see only tiny verticals being tried by OpenAI and Anthropic, with Claude Design et al

This can be better done by other companies, with a fresh balance sheet, don't you think?

There'll be plenty of people at Anthropic and OpenAI opening those companies

Nowadays you can just start with a plan.md and add everything you want there, verification etc

And you can build a loop to verify/validate it in case its a big/long-running plan

I don't have the problems you mentioned that the models aren't able to follow instructions throughoutly, they do make some mistakes, but it's up to you to set up a process that would work for your codebase to understand that the mistakes were made and need to be fixed, that can be done in a loop for example.

DeepSeek is model that you can run anywhere, meaning that any datacenter/company could run it at a price that would give them a margin taking in account running costs of the datacenter + depreciation.

I'm very sure we're overbuilding capacity, and we'll find out around 2027.

Those open weights will become extremely attractive because all that datacenter extra capacity laying idle will be possible to be used by builders.

It may be the case that in 2027-2028 will be cheaper to start a new AI model company from scratch to compete with Anthropic/OpenAI as you'll be able to:

- the new company won't be tied to so much capital, debt and running costs

- get much cheaper datacenters to train than it is Today, as the big labs will be stuck with bad infra contracts

- have great engineers/researchers willing to jump ship because if Anthropic/OpenAI valuation falls, that will destroy their equity

- wide available chinese models to distill from, papers and ideas on how to acquire datasets

Anthropic/OpenAI are playing an extremely risky game, they need to greatly keep doubling their bets and delivering model big performance increments.

If they don't give us ASI/AGI in 2 years or find huge new markets, there's no way the economics will make sense.

The execs forgot why AI models even exist (like huge usage for the cyber models, or new types of models)...

AI models' usage is mostly tied to build Software Today, that's where the value really lies now.

If there isn't enough Software to be built as people claim to the rate they expect, eventually the supply of AI models will outstrip the demand.

You can't sell pick & shovels infinitely, AI models themselves are also in that category.

Eventually you need enough holes to be dug, and gold to be struck, and nobody knows what's that demand. (even though they built all that supply...)

Context, similar to your computer's memory, is limited.

Additionally, the more things the AI has in context, the worse it performs.

Skills helps you to quickly add more context for doing things, without polluting your context window.

I think you need to learn about AI, there isn't anything hype there. It's just an easier way of giving more context to the AI.

This can be helpful, for example, if you have an internal API that does XYZ, and you want the AI to know that API, you can save that API, its endpoints and authenticaiton methods to a skill, and invoke it when you want your harness (CC, Codex) to know about the existance of that API, to make requests, make tests etc.

If good, continue with it, if bad, give up.

Demos are so good for that purpose, if you show the game loop and it sucks, you'll need plenty of work to change this situation around (if even), so try it.

Also about privacy-first apps.

People have no idea and everybody pretends to be an expert and ignore how good China is on AI research

It may not often use them often as it uses software, or it may not remember all integrals rule tables, but it understands the concept, it's usefulness, and in case he needs it, he can figure out a way to work it out.

This is the same for Software Engineering and code. You don't need to have written the HTTP protocol library in order to use it.

I haven't written a single line of code since more than an year, but I've made the AI write multiple thousands of lines of code since then.

It's just a calculator. You still need to know how to use the calculator and for what purpose do you use it.

What changed is that you no longer will need to learn all ins and outs of coding.

I have a hard time understanding this.

We have plenty of adults with terrible social media addiction that is destroying their lives, and nothing being done about it.

For example, for the US to have a chance in the EU, it would first need to fix its YOLO fiscal policy of sustained 5.5% debt/gdp deficits.

We shall see in a few years as US's debt balloons and the average American becomes pseudo-slaves from a few overlords... to see if the EU is really bad as some Americans believe it to be.

I love working with AI and agents.

But I don't know how long this will go, so I'm investing on getting fit again for random jobs like moving atoms.

Or realizing my life's dream: become a lifeguard in public swimming pool or learn the trades to maintain swimming pools. There's a shortage of them where I live and it's enough money to pay for food etc.

I love swimming and watch people happy enjoying the water, it's such good vibes I'd do this forever.

Our life is mostly suffering and grinding through... but people on swimming pools are always smiling, having a good time & having the basic pleasures of enjoying the water and the moment.

Blizzcon canceled. All of its IP barely got any love.

See what players think about the latest World of Warcraft patch. It's absolutely shit and broken. People say they fired the entire QA department since a few years back and since then the quality has just gone down.

They buy those businesses because they have nothing to do with that free cash flow, and for accounting reasons it makes sense to have them.

They didn't buy those businesses to develop it further and make it worth more.

Github will just become ever more irrelevant.

The key issue is that the US governments let those huge monopolies exist, and then use their money to buy other businesses and enshiftify them.

Unless that changes in the US, this will continue happening.

LLMs for language feels like it's definitely the way to go. I feel like that by just improving it further can definitely reach perfection, if not very close.

My concern is mostly all adjacent fields, like systems thinking, spatial reasoning, "real" human-like reasoning etc or as you put it, "AGI".

Doesn't seen this will take us there at all. I don't feel like we're closer to AGI than we were on the earliest versions of ChatGPT.

MoE: I assume some people just specialize in working with routing as with that, as by reducing the amount of params and just using a subset, you end up making it less costly. So, AI researchers are only working on optimizations on getting this better?

Same question on Reasoning, so AI researchers are working mostly on optimizations on top of it, like CoT and so on, like mini-optimizations.

So basically, they work on those micro-optimizations, put them together and see a % improvement in a benchmark?

I'm sure this is probably awesome for languages, which if I'm not mistaken, it was the use-case initially used on "All you need is attention" and the entire LLM revolution.

But this seems to be a very clear path to be "taking the car to the carwash by foot" for a long time, isn't it?

It feels like we'll keep "taking the car to the carwash by foot" until somebody optimizes for that prompt, or some pre-training done, and then there'll be another prompt that will show that the AI has real trouble with very basic real-world reasoning and imagination.

Isn't it the case, or do you see any kind of research that could take us from that plateau full of micro-optimizations that get us a few cm higher to the peak?

No shop will want to have the expectations to successfully fix a complex GPU, unless it's their main business, as the solution can potentially not be 'reballing the memory modules' and they might just break your GPU in the process, they don't wanna have to pay you back in case they do, otherwise the risk they are undertaking isn't worth the chance of they fixing it and getting your X euros.

Go to hardware communities and ask them if they'd like to have a go at it, find somebody reputable and go with them.

Otherwise based on what you said, you literally will have to trash it, or find a way to constantly cool it down in a very reliable way (Good luck! I tried the same with a CPU with an overheating issue and failed miserably. Hopefully you'll manage it)