Now there's a security researcher I can't imagine having confidence in...

If it was a toy inbox, ok, good thing to play with, but on an actual inbox, with the universally recognized badness of OpenClaw, and a *security* engineer... Not even a misguided software person that just doesn't take security seriously enough which is bad enough, but someone who by any vague measure *should* know better...

So the hypothetical is that entry level coding jobs are toast but you still need the advanced folks to actually direct things, and they will need to be able to review and amend code.

In such a scenario, then education takes over the role of 'doing stuff humans don't have to do anymore, but still need to know how to do it'. Like how math education starts by banning calculators, then as the education advances increasingly advanced calculators and computer software are allowed to handle the tedium that was left behind.

Note that white collar jobs aren't all 'intelligent', there's a fair amount of tedious manipulation of purely abstracted data in computers. This is the part that, in theory, maybe, could be changed by LLM approaches. Some debate can be had about which white collar jobs are which, and how far the LLM can go or not, but those at least are in the ballpark.

If a part of a given blue collar job couldn't be done with gloves more substantial than medical gloves then it's pretty far from being within the reach of any AI technology today. The approaches to do anything require training data, and instrumenting the dexterous and sensitive work of hands is just not a thing. It's why the humanoid robotics demonstrations remain embarrassingly bad for these companies that *usually* at least can pull off superficially impressive demos. The fact that even without AI, the tele-operated demonstrations fail because humans suck at doing this stuff with even gloves on, let one trying to do it remotely with controllers.

By comparison, driving is just impossibly easier for machine learning to try to take a swing at. Humans operating almost entirely based on vision and input entirely into two pedals and a wheel. Supremely instrumented and very coarse grained controls to manipulate. Even with many lifetimes of "experience" to work with for relatively simple actuation, this has been a huge challenge and still not closed. HVAC and plumbing are so much more complex with inputs that aren't possible to capture as training fodder, with lower possible volume of training to get even if you could instrument everything.

The strange thing is the bubble can burst at the same time as it can also disrupt white collar labor dynamics, regardless of whether it is capable or not.

It's no secret that office labor force has a great deal of padding, and a shift in business leadership mindset could cause a reckoning. Generally office workers are presumed to be worth keeping around for various non-core reasons and if 'AI leadership' displaces the ego of 'I have lots of professionals under me' as the valued attribute, well they might start measuring themselves by how many people they dispose of rather than how many people they have authority over.

Things is that folks already know perhaps most office workers are superfluous. AI won't shine a new light on that, though it could rationalize some shift in strategy towards them.

Several factors into why office workforce is padded, people wanting to brag about number of professional reports, people wanting sacrificial headcount for layoffs, leaders having no idea who is and is not useful, politics sometimes outclassing productivity...

I think it says a lot about the amount of uncertainty in the market. Nobody knows what's going to happen next, and investors are skittish.

It's just what Kurzweil predicted -- as the pace of change keeps accelerating, peoples' ability to predict the future gets compressed into a shorter and shorter time-window, like driving at night and outrunning the headlights of your car.

They used to make really good plasma sets back in the day.

Scripts are usually poor performance for the fork/exec that you usually do.

The scripts get really scary when they manage to avoid the fork/exec subcommands and manage to do almost everything in the script primitives themselves.

Would much rather see a perl or python script than the monstrosities I have seen...

Though I will confess conversely I hate seeing hundreds of lines of python that could have been a quick readable 6 line script (performance permitting, as you point out).

The technical challenge of porting COBOL has never been the impediment. It's the "no one wants to own replacing code with almost no 'glory' and all 'risk'".

LLM may very well be capable of modernizing COBOL, it's plausible. It's still risk without particular potential for enriching glory for their trouble.

Comments only do so much, the syntax for shell scripts to do complex things just gets ugly.

Shell syntax is optimized for running commands, but that leaves fitting other tasks into very obtuse syntax. You also have no sane way of managing namespace. It is *capable* of a lot, but other languages are just more sophisticated at scaling complexity, at the expense of being super tedious for "just run some commands". LLM offloads tedium so a 'script-level' description can be formed into a more structured program, and potentially more resistant to things like shell injection.

Scripting is nice and the best when you have something simple to do. But it just starts to suck when you have something not so simple to do.

Particularly interesting is how the assumption is that for this round of tech, the successful LLM companies will succeed in avoiding any on-premise competition and thus the datacenters are obviously a long-term requirement.

This is a huge 'wtf' statement. People use shell scripting because it is low friction, but shell scripting sucks for scaling up complexity and maintenance. If you said to me 'LLMs let people take their quick and dirty shell scripts and port them to something that has more reasonable syntax for complex structures', ok, that sounds plausible and a good idea.

If a shell script is so complex you *need* an LLM to do it for you, it *probably* shouldn't be a shell script.

(Frustrated due to a certain team I've worked with refusing to move beyond shell scripting and having just the most unmanageable ball of crap, and they frequently have to ask others to debug the mess they made because they no longer understand their own scripting)

That's been the bigger problem for them. Fortnite and Roblox thrive on pop culture tie ins and creative freedom. Meta has largely gone the PlayStation home approach where only what they curate is there, but even worse as Sony had a bit more creativity at least.

You basically have Nintendo Miis running around bland environments with nothing to do or look at. Nintendo Miis got boring over decade ago but Meta still seems hung up on the concept..

Breaking the vr dependency is unlikely to do anything for the concept.

Going forward, authenticity is going to be a rare and therefore valuable commodity.

The platform that figures out how to maintain a user base of real, sincere, honest human beings will have an advantage over its competitors that are nothing more than a raging sea of ads, trolls, bots, and AI slop with the occasional drowning human mixed in but on his way to the exit.

I'm not sure what the formula is (if I knew I'd probably be rich), but maybe something combining credit checks, public/private key identity authentication, and a reputation system that people care about maintaining?

That the prevailing winners in the market are not winning based on code, but on marketing. Now amount of 'here's a cheaper knock-off' is going to overcome the scare factor. Saving money won't win any of these decision makers enough to warrant the risk of cocking up something important.

However, AI is pitching supreme bespokeness. The cover is not just about 'saving money', the emphasis is that it can be fit to your 'special' purpose.

It's all a marketing game, and India low cost labor didn't bring any marketing to the table, but the entire damn world is doing marketing for the LLM bros.