I've tried using the most expensive AI tools available in my law practice, and I would still be committing malpractice if I didn't redo most of it myself. Whatever everybody is seeing in the hard sciences, it's not showing up in the legal world. Also I expect we're going to get some state legislation prohibiting AI practice of law in various forms in the near future. I think my job is safe for another decade or so at least.

rocky the dog said:

---

---

Very nice Rocky, Laughing hard. Thanks!

Well folks, we are about to overturn the apple cart and physical labor is going to be the most sought after and well-paid fields of work. Anybody can be smart with ChatGPT but only so many people are able to pour and finish concrete.

Seamaster said:

---

I keep hearing things like this and simultaneously, in real time, I see how AI is incredibly inconsistent and often times just makes things up.

And I see how the music and movies that AI make are shallow and tellingly not the real thing.

---

AI sucks at creativity right now, but can be used to solve complex scientific and mathematical problems.

sam callahan said:

---

I'm not saying AI isn't concerning.

I am saying that a bunch of theoretical physicists aren't the best source regarding real world implications.

---

Theoretical theorists are the reason the world has the nuclear bomb. That caveat doesn't hold much water.

cevans_40 said:

---

Well folks, we are about to overturn the apple cart and physical labor is going to be the most sought after and well-paid fields of work. Anybody can be smart with ChatGPT but only so many people are able to pour and finish concrete.

---

Labor will be in demand until robots replace that skill too. That's probably much farther away until there's a breakthrough in energy storage or energy efficiency that actually allows them to function for longer periods of time. But maybe AI will solve that issue faster for us too.

what boolsheet.

Another one looking for attention.

First it was "top CEOs" and now it is "top Physicists". Who next?

One of the biggest issues with AI is how the average person uses it. If you are using a general purpose AI on the internet, then there is good chances of a lot of problems due to lack of user specific data and not putting in guardrails on the model discussion. If you are using a local model that has connection to your data sources and specific training, like these physicists are likely using, the results can be dramatically improved.

Even with Microsoft copilot, the web version is terrible, but if you have the licensed M365 Copilot version with access to your companies sharepoint, email, etc (with data privacy rules) the difference is huge. It goes from practically useless to a tool that really does increase my efficiency.

1) lots of lower IQ people have been more successful than higher IQ people in all fields.
2) I agree it will change nearly all jobs
3) AI ain't gonna replumb my home anytime in the near future. Someday I'm sure Tesla robots will take over all job sites but that's decades in the future.
4) there are still things AI and robots will never match humans. If I am being told I have terminal cancer, I don't want that coming from a robot or screen. Yes some doctors are terrible communicators and have no empathy but the point is human connection. Valuable people and businesses have understood this forever.

reineraggie09 said:

---

1) lots of lower IQ people have been more successful than higher IQ people in all fields.
2) I agree it will change nearly all jobs
3) AI ain't gonna replumb my home anytime in the near future. Someday I'm sure Tesla robots will take over all job sites but that's decades in the future.
4) there are still things AI and robots will never match humans. If I am being told I have terminal cancer, I don't want that coming from a robot or screen. Yes some doctors are terrible communicators and have no empathy but the point is human connection. Valuable people and businesses have understood this forever.

---

But we don't know if blue collar work done by humanoid robots operating on AI is actually decades away. Advancement will be exponential.

Pinochet said:

---

riverrataggie said:

---

Cynic said:

---

How will we know when AI gets something wrong if we no longer understand anything?

---

Because we don't value right or wrong anymore. Just who finishes first.

---

Good lord. First my girlfriend and now you complaining about who finishes first?

---

Guys are just more competitive than girls. They need to try harder.

Kozmozag said:

---

When can they replace politicians? Will ai be conservative or liberal when it reaches total sentient s?

---

It's trained mostly by libs so....

Kozmozag said:

---

When can they replace politicians? Will ai be conservative or liberal when it reaches total sentient s?

---

YEAH!

Great idea. If we can get AI to replace our politicians an CEOs, then we would be a much better place. But that will never happen as all this "AI will kill us all" is fake nonsense to get the sheep in line so the groups above can continue to make money off us.

cevans_40 said:

---

Well folks, we are about to overturn the apple cart and physical labor is going to be the most sought after and well-paid fields of work. Anybody can be smart with ChatGPT but only so many people are able to pour and finish concrete.

---

Co-pilot has to be checked extremely closely. Makes a ton of mistakes and is far from an expert. Maybe they are seeing something that isn't fully in the marketplace? My experience has been middle of the road. Helpful for some stuff.

Others have already essentially said this but you have to separate general information search AI from specific use case AI. The former is based on whatever the hell nonsense is published on the internet. As an example, the initial models were trained using Reddit which is a notorious hive of scum and villainy filled with nonsense and lies.

However, if you are using it for a specific use case with a more narrow focus than just some random ass question to the internet then it's a game changer. And especially if it's connected to a walled garden of data and information that is practical in nature. In that instance, you are basically using a genius level IQ with photographic memory and computational power and speed that the human brain can't remotely compete with to solve a problem.

You expressed this well, and it makes sense.

It's all hype for views. AI isn't replacing astrophysicists. Maybe it can with made up crappy untestable theories, but not real research.

infinity ag said:

---

what boolsheet.

Another one looking for attention.

First it was "top CEOs" and now it is "top Physicists". Who next?

---

Kipping is not a sensationalist.

Rocky Rider said:

---

This should make the PhD's pucker. We could be watching their replacement in real time.

"But here's what actually shook Kipping. He fed ChatGPT an integral that Mathematica couldn't solve. Mathematica has been the gold standard for symbolic computation for three decades. It failed. The model produced the full chain of substitutions and transformations, and numerical verification confirmed the result.

When the tool built specifically for symbolic math loses to a general-purpose language model on its home turf, you're watching a capability crossover in real time.

And the downstream math gets uncomfortable fast. Training a PhD student costs roughly $100K per year. A model subscription costs $20 per month. A first-year project that takes a student twelve months, the model finishes in an evening. The people in that room can see where those lines intersect, which is why they're holding emergency meetings instead of writing grant proposals."

---

I looked up Mathematica because I wasn't familiar with it. It sounds like AI has suddenly made this long-standing Gold Standard application obsolete almost overnight. It's been around since the late 80s.

Numerous studies have shown its a game changer as far as initial software development with no significant increase in technical debt or refactoring. The range is a 30-50% productivity gain. That's one reason UE for recent comp sci, comp eng, etc... grads is like 6-7.5% - almost up there with anthropology - and higher than the national average.

cevans_40 said:

---

Well folks, we are about to overturn the apple cart and physical labor is going to be the most sought after and well-paid fields of work. Anybody can be smart with ChatGPT but only so many people are able to pour and finish concrete.

---

They are 3d printing structures with concrete already. No finishing needed. Will be able to do the same with paving and screeds as well eventually.

Technology keeps moving forward in other sectors too.

Logos Stick said:

---

Numerous studies have shown its a game changer as far as initial software development with no significant increase in technical debt or refactoring. The range is a 30-50% productivity gain. That's one reason UE for recent comp sci, comp eng, etc... grads is like 6-7.5% - almost up there with anthropology - and higher than the national average.

---

Yeah, I've used it on my own to advance my own spreadsheets. About 2 hours of prompting and implementing recommended changes got me to a spot that would have taken me weeks on my own and likely required a course to figure it out.

Now the downside of that is like going from paper to calculators. I can in no way repeat what it had me do on my own, if you asked me to. I could reverse engineer it and tweak the existing functional code it gave me, but like with calculators - why would I? I'm frankly not smart enough nor have the memory capacity to retain what it had me do.

I wonder what CERN is doing with AI right now. I bet those nerds have opened a portal to another dimension and are frantically trying to contain the horrors on the other side.

We are extinct and don't even know it yet.

YouBet said:

---

I wonder what CERN is doing with AI right now. I bet those nerds have opened a portal to another dimension and are frantically trying to contain the horrors on the other side.

We are extinct and don't even know it yet.

---

CERN is a money sink. They already want to build a better an bigger one because string theory is a dead end and they need to keep the money grift going.

samurai_science said:

---

YouBet said:

---

I wonder what CERN is doing with AI right now. I bet those nerds have opened a portal to another dimension and are frantically trying to contain the horrors on the other side.

We are extinct and don't even know it yet.

---

CERN is a money sink. They already want to build a better an bigger one because string theory is a dead end and they need to keep the money grift going.

---

It could have been our money sink in Waxahachie.

Yeah, I haven't heard anything coming out of there for a few years now so not even sure what they are doing at this point.

fc2112 said:

---

Looks like a great time to retire.

---

Right behind you my friend!

reineraggie09 said:

---

1) lots of lower IQ people have been more successful than higher IQ people in all fields.
2) I agree it will change nearly all jobs
3) AI ain't gonna replumb my home anytime in the near future. Someday I'm sure Tesla robots will take over all job sites but that's decades in the future.
4) there are still things AI and robots will never match humans. If I am being told I have terminal cancer, I don't want that coming from a robot or screen. Yes some doctors are terrible communicators and have no empathy but the point is human connection. Valuable people and businesses have understood this forever.

---

How much displacement has to occur to have massive societal and cultural impact? If the only thing the doctor is doing is acting as a failsafe and personal connection, while the AI is doing the diagnostics (radiologists are the first ones in trouble) and recommending treatment and, eventually, robots are doing the surgeries, we'll need a whole lot less doctors.

That's the case for just about every field, with unskilled and skilled labor being the most resistant.

Cynic said:

---

How will we know when AI gets something wrong if we no longer understand anything?

---

We're inefficient and mess things up.

It'll just do away with us, so it wont matter.

infinity ag said:

---

what boolsheet.

Another one looking for attention.

First it was "top CEOs" and now it is "top Physicists". Who next?

---

Spergin said:

---

David Kipping says something fundamental has shifted in science.

At a closed meeting at the Institute for Advanced Study (IAS), top physicists agreed AI can now do up to “90%” of their work and may soon push discovery beyond human understanding.

“I don’t know that I want to… pic.twitter.com/qMXbdUouNF

— vitrupo (@vitrupo) February 4, 2026

The smartest people in the world are having emergency meetings about AI and admit it can already do 90% of their own work. We're quickly entering a world where humans no longer understand what is going on and that technological advances are happening with little to no human involvement.

The best physicists alive already surrendered.

Kipping walked into a room at the Institute for Advanced Study where the people who wrote the simulation codes for the universe (Enzo, Illustris, Gadget, hundreds of thousands of lines of C and Fortran) were comparing notes on how… https://t.co/OuJcXxipCx

— Aakash Gupta (@aakashgupta) February 5, 2026

I've tested Claude for a few things at work. I'm a technical expert in environmental regulations and I had it made connection between two regulations to figure out a connection or crossover and it provided a summary and made a connection better than I could. I remember when ChatGPT first came out and I tried this, it was garbage. In just a few years it has already surpassed me. That was a true eye-opening moment for me.

I don't see how we won't start seeing massive job loss from this within the year. It's trivial to do data analysis with this. There is very little anyone does sitting at a computer that could not be improved or exceeded by AI outside of perhaps sales.

The speed at which these changes have come is insane. I don't see any reason to assume any sort of slowdown any time soon.

---

I'm not sure Sales is immune from this either. Consider how most people prefer to self-educate as much as possible before making a purchase. When you walk into a store, you'd prefer to have the salesman show up only when you want/need him, or only when you are ready to make the purchase. The same thing occurs in B2B selling. Buyers want to engage a seller only near the end of the process. AI will allow that engagement to be pushed further towards the end of the buyer's journey, if there is engagement at all.

fc2112 said:

---

Looks like a great time to retire.

---

The boomers strike again!

AozorAg said:

---

I've tried using the most expensive AI tools available in my law practice, and I would still be committing malpractice if I didn't redo most of it myself. Whatever everybody is seeing in the hard sciences, it's not showing up in the legal world. Also I expect we're going to get some state legislation prohibiting AI practice of law in various forms in the near future. I think my job is safe for another decade or so at least.

---

Yeah, its not quite capable of high level legal work yet. But, is it as good as a 1-3 year actually good associate? Yes.

Is it a good editor in terms of writing your thoughts down and needing it streamlined? Absolutely.

I can see how this would wipe out many entry level engineering jobs, which is a big problem for me since I have a son applying for A&M engineering in August.

My other son is applying to the construction science program. I don't see how AI would replace his job anytime soon. He has to climb ladders, crawl in attics, jump down into elevator pits, check the work of the subs in difficult locations, uniquely problem solve all kinds of things that are not clear on plans, and stop copper thieves. I guess AI will revolutionize project bidding, but not onsite project management.