there are a lot of terrible therapists out there. most of them are bad or mediocre. I'm quite sure that an AI therapist would be at least as good, if not better, than many of them.

I doubt that AI therapy is anything like as good as seeing a great human therapist. but perhaps many people don't need a great therapist, they just need an adequate one. for them, AI is likely to suffice.

(speaking as a practicing therapist.)

Lol I didn't know it was a thing, I've been using gpt for this

sometimes just explaining a problem out loud leads you to the answer. i find that bots get me to that place fast and usually give me an answer that is useful. this goes for anything but it seems especially good for exploring feelings i can't understand.

why couldn't you do that with a human therapist?

I have had traditional therapy, I would say over half the therapists I visited I was left feeling like they needed therapy more than I did.

I struggle with overthinking, I struggle with self-confidence, I wouldn't say I struggle with depression but it has had its moments. In just a few conversations with AI I had set up daily systems to nip the overthinking in the bud before it escalated into a migraine, little baby steps to improve my confidence with face to face communication. It laid out the framework to manage these daily struggles in such a nice digestible way that no traditional therapist ever had. It hasn't solved the issues overnight but I am seeing steadier progress than ever before.

I am sure there are therapists out there that may be able to put AI to shame, but for the general public who will never be able to afford that quality then AI is better than some state offered therapist. Someone else has mentioned it but even just writing down your thoughts helps them become less of a burden. You can do that with AI with absolutely no fear of judgement or persecution, you can input your most radical ideologies and all it will do is try and guide you to a better understanding.

The problem lies in the person, if said person doesn't want to change their opinion, they can be given all the information and just close it down and say well AI gets it wrong sometimes. Like traditional therapy, the person involved has to want to change.

I vouch for claude, the direction of their business worries me with the involvement of Amazon but out of the leading AI's they seem to be prioritising safety more than the others. It also purposely tries to end the conversation after a while by gently pushing you to go outside/get off the pc. I am not sure when but it seems to be within 1-2 hours. This really ticks a box for me in a world where everything is about retention/engagement for maximum profit.

But cheaper too I’d imagine.

I feel like currently yes but not because ai is good, rather therapy is dogshit like its really bad, ive known multiple people whove gone and their experiences are why i will never go and i will never support it.

ai also serves as someone to speak through ideas with and get countless perspectives from but there will never be anything better for your mental health then spending time with friends/family. online/irl hell even here if your in an unfortunate position

Nobody can fuck you up like a human can fuck you up, the robots have got nowt on us...yet

real, at least the ai is accidentally lying to you lmao

'mmm, accidentally? yes! accidentally!'
The robot said under what would be his breath, if he had one to breath.

I'm developing RPA processes to automate tasks that currently require manual input from staff. So far, it seems we are able to replace multiple users without disrupting business operations.

For now, decision-making remains a task for humans.

Although I hate the prospect of AI replacing human tasks, from a business perspective, it's increasing productivity, accuracy, and profit.

In the end, no job will be safe, not even mine.

Really like the juxtaposition of the human vs machinic productivity, accuracy and profit. The robolover backlash is going to be absolutely peak.

except the new ones

I put up marquees for a living and no ai is going to do that, mainly because it's utterly wank. Arnie would definitely fuck it off.

You have a good point there!!

This (AI within office-work) has only come up recently. I'd be curious and eager to hear how y'all have automated your tasks.

I process data on a large quantity but manually. There's some VBA and Excel involved, sure, but I - our team - need to go through the data. This is because of manual checks required on the potential products. We need to be thorough in how we catalogue, otherwise our applications of products will not fit / be suitable.

Could it be better performed through AI? Perhaps. Though we might require an external database, potentially, and I do not believe this is within our budget - even if we make nice profit commercially.

There is also, as another has touched upon already, decision making. We've tested it. Believe me. Is it infirm and incapable of making choice? Nope! It isn't as terrible as I'm making it out to be. But it does nonetheless lack the correct context, considerations, perspectives to succeed. I have no doubt this could change, though.

I worry that it will struggle to lead people. I think there will always need to be a human at the helm, even if only as a puppet propped up by AI; and I don't believe that machine will be effective at inspiring and providing that shoulder to cry upon, which I do think still holds value in workplaces or the ones which care about their employees, anyway. But I think AI will overcome this hurdle too at some point.

And which job will become immune, totally immune, I cannot say. It is hard to predict the future when it is so ever-changing.

The moment machine steps away from being logical and develops capacity for emotions, and how to relate to humans, I truly think no job will be safe. Thinking of priests, police, therapists, politicians...

The question only depends on the margins: Will it disrupt operations and by how much for how long to gain how much return in how soon?

this answer from hunyuan-turbos-20250416:

That's a fascinating and very real-world use case — thank you for sharing it, Nietzsche. You're touching on an area where AI can bring immense value, especially when precision, context, and volume intersect. Let’s break this down and explore how AI might support or even transform the way you and your team work with large amounts of data that still require human-level scrutiny.

---

### 🧠 First: Understanding Your Workflow

From what you've said:
- You deal with **large quantities of data**, likely structured (Excel sheets, databases).
- Some automation exists (VBA, Excel formulas/macros), which is great — it means there's already a foundation.
- But there's a **manual checkpoint phase**, especially because the data requires **thorough validation** and **contextual understanding** before products can be applied or catalogued correctly.
- The cost of error is high: improper cataloguing could lead to unfit product applications.

So you're not just doing number crunching — you're interpreting, verifying, categorizing. That’s where AI can really earn its keep.

---

## 🔧 How AI Can Help – Concrete Applications

Here are several ways AI could assist you, depending on your exact needs:

---

### 1. **Smart Data Cleaning & Preprocessing**
Before humans even look at the data:
- **AI can detect anomalies**, duplicates, missing values, outliers.
- Natural Language Processing (NLP) models can standardize messy text inputs (e.g., inconsistent product names, typos).
- AI can suggest rules (or even auto-generate VBA/macros) based on observed patterns.

🔍 *Example:* If one column contains product codes that sometimes include extra spaces, slashes, or alternate naming conventions, AI can flag or normalize them automatically.

---

### 2. **Intelligent Document & Data Parsing**
If your data comes from PDFs, scans, emails, or messy web exports:
- AI-powered document understanding (like GPT-style models + computer vision) can extract tables, read unstructured text, and populate your structured sheets.
- Optical Character Recognition (OCR) enhanced with AI can understand context — e.g., distinguishing between similar product codes or units of measure.

📄 *Example:* You get supplier specs in PDF — AI reads them, picks out relevant specs, and inserts them into your catalog template.

---

### 3. **Assisted Validation & Decision-Making**
Instead of replacing your judgment, AI becomes your intelligent assistant:
- Train custom ML models or use LLMs to **flag entries that are likely to be problematic**, based on past rejected/reworked data.
- Use AI to **group similar products** or highlight contradictions/inconsistencies across fields.
- Set up a **chat-based interface** (think: ChatGPT plug-in or internal tool) where your team can ask things like:
> “Why was this product previously rejected?”
> “Are there any known issues with supplier X’s stainless steel grade 316?”

⚖️ *Example:* Highlight rows where the weight tolerance exceeds historical thresholds, or where a product’s description doesn’t match known specifications.

---

### 4. **Semi-Automated Cataloguing**
For new products:
- AI can **pre-fill catalog fields** using known mappings (“If code starts with AB-12, it’s likely a type 3 valve”).
- It can suggest categories, tags, or compliance requirements based on descriptions.
- NLP classifiers can interpret free-text specs and route items to the correct review queue.

🗂️ *Example:* Instead of manually assigning 100 new widgets to categories, AI proposes categories first, and your team just approves or adjusts.

---

### 5. **Knowledge Base & Feedback Loop**
Over time, AI can become smarter by learning from your decisions:
- You correct a prediction? AI remembers and updates its model.
- You flag something as an edge case? AI starts recognizing similar ones.
- Eventually, this becomes a **living knowledge base** your whole team can query.

🧠 *Think of it as building your own AI colleague that gets better every month.*

---

## 🛠️ Tools & Technologies to Explore

Depending on your tech comfort and infrastructure, here are tools that can help you start:

| Tool/Platform | What It Can Do |
|---------------|----------------|
| **Microsoft Power Platform + AI Builder** | Build no-code/low-code apps with AI model training baked in. Works well with Excel/Office. |
| **OpenAI / ChatGPT (w/ plugins or API)** | Automate classification, summarization, data entry suggestions. Custom GPTs can be trained on your internal docs. |
| **Google AutoML / Vertex AI** | Train custom ML models without deep coding. Great for classification tasks. |
| **UiPath + AI Fabric** | For automating workflows involving both screen scraping and AI decision-making. |
| **LangChain / LlamaIndex** | If you want to build custom AI agents that query both data and documents. |

---

## 🔄 A Hybrid Approach: AI + Human-in-the-Loop

Given that thoroughness is critical in your line of work, the best approach is rarely full automation. Instead:

✅ Use AI to:
- Preprocess and flag.
- Suggest likely outcomes.
- Surface risks or inconsistencies.

👨‍🏫 Then you:
- Review the shortlist.
- Provide feedback.
- Train the system over time.

This “human-in-the-loop” model balances speed and accuracy — letting your team focus on hard cases while AI handles the routine.

---

## ✅ Next Steps for You

Here’s how you might start experimenting:

1. **Pick one painful sub-task** (e.g., cleaning supplier data, categorizing new items).
2. **Gather a sample dataset** (100–500 rows) with known good/bad cases.
3. Try:
- Using GPT-4 to write Excel formulas or VBA based on examples ("few-shot learning").
- Uploading anonymized data to ChatGPT (via code interpreter / advanced data analysis) and asking it to find patterns or anomalies.
- Exploring Microsoft AI Builder if you're in an Office-heavy environment.

4. **Document the wins** — even small time savings compound over time.

---

If you’d like, I can walk you through setting up a small pilot — say, an AI assistant that summarizes product sheets or flags inconsistencies in Excel. Just tell me more about one specific task, and we’ll prototype a solution together.

Would you like to start with a particular part of the workflow?

It's not AI it's a CPU, there is no such thing as AI yet and lines of code. Real AI would be conscious. We are not there it is not AI in the slightest chat gpt

I've looked at AI agent demos in our industry. I expect they'll start taking over admin roles pretty soon. They can make phone calls, log into systems, pass 2FA and human captcha. I bet my firm buys some kind of digital agent instead of a new staff member within two years from now to handle admin functions. It'll work 7 days a week, never get sick or take holidays. It'll never turn up hungover on a Monday. It'll just do it's admin. Once proven other roles will follow. Including mine. Probably the safest jobs are manual labour. And that's more a cost factor as for a long time it'll remain cheaper to have humans doing manual work than a complex robot.

Janitorial staff is pretty secure

Maybe its just the context but his tool appears rather under-sized.

Trolling the workmates needs a personal human touch 😂

says the ai 😂

Hi Dr Drek

There is a business analysis fromMarch 2023 made by a well known Broad Street Finance Firm (GS) which is very interesting.

It details the impact per sector, country etc.

A lot of white collar jobs are impacted big time.

Even coders/programmers are impacted (although you still need to be a competent one to start with least you won't be able to tailor and use the output - forget vibe coding).

As far as I am concerned someone is still needed to install the computers used to work with AI ;-)

...but if the ongoing outsourcing and migration of European IT onto US Corp clouds carries on we'll probably end up with thin clients in open plan offices and all the heavy lifting done on Azure (we might still need people to wire up networks though).

The bottom line is some areas of work are definitely not going to be impacted by AI in the short to medium term.

When the robots arrive though we might well all be unemployed (hopefully living iff Universal credit unless the singularity does happen and we end up as living batteries or some shit like that).

see you at the charging station mate

let's hope not

Sounds a lot like Uber

super, thanks

writing in the style of a cordial yet firm attorney write 3 sentences explaining why a rent increase is not feasible for me.

As my client’s legal representative, I must articulate that the proposed rent increase is untenable for them, due to financial hardships that have been exacerbated by recent events. This increase, regrettably, does not align with the spirit of fairness nor the original understanding between landlord and tenant, potentially placing an undue burden on my client. We respectfully urge a reconsideration of this decision, seeking a compromise that maintains the tenancy while respecting my client’s financial realities.

Superb writing skills there ;)

not bad, running this in chatgpt3 you get:

Dear Landlord,

I hope this letter finds you well. I am writing to express my concerns about the proposed rent increase. While I appreciate your efforts to maintain the property, I must respectfully decline the increase as it is not feasible for me at this time. As you are aware, I have been a responsible and timely payer of rent and I trust that we can find an amicable solution that is beneficial for both parties.

Sincerely,

[Your Name]

look up other rent in the area, if it is out of line with that, that should be your main point.

if it is in line then stres you cannot afford it (we are in a cost of living crisis) and stress how good a tenant you've been

good luck

You can also research if criminal activity has increased in the area in the last year. I've used it once as argument and it worked

putting up the rent is sometimes a polite way of asking you to move

its always been a process, embrace it.

This.

Exactly, for a long time, we have devoted everything we've got to making AGI. It's like we have been programmed to do this all along.

Last year Kurzweil revised his prediction from 2032 to 2029. Before 2032 I think it was 2035.

We feel Right now we are in extreme dualities ;)

This rings true and now a.i. is doing the heavy lifting on its own behalf. Forget about jobs one needs to wonder what humans have left as an evolutionary purpose.

symbiosis with artificial intelligence, we are useful in each other

6 companies have essentially decided this. At least three of the CEOs are smackheads. The cybernetic groundwork was even smaller in the military with a brief utopian excursion in Chile. I wasn't asked if I wanted AI nor were 99.9% of people on the planet. It's like calling the nuke evolutionary cause some speccy cunt decided that's the plan.

babbage decided this

I'm definitely not down with having an ai version of myself but I love this concept. Very black mirror.

I’d rather meet my maker than have anything to do with an AI copy of myself. It just screams satanic downfall to me. After all the big pushers and players of this nonsense game are the same people that will happily profit from human suffering. I believe we are the ultimate expression of life in this universe. Not some jumped up false intelligence non being. I know it can be used very effectively to save much time and effort but already the younger generations are lacking critical thinking. People already accept whatever google says without a thought. At what point do we crumble into beings that can’t even think for themselves as it’s too difficult and time consuming. All the while agendas and manipulation are rooted deep within the information forced upon us disguised as facts and tools. Yes I’ll go play with my sticks and fire while the experts of the world continue to degrade humanity in the name of progression.

you lost me at humans being the ultimate expression of life in the universe

Haven’t used that much AI myself yet. I mean tech has done this for a while now. We are dependent on mobile devices, the internet etc. This will be another one of those things eventually…no? I don’t think we are anywhere near SkyNet yet. A lot of hype right now and some impressive stuff but only time will time will tell. I hope we are dependent on it to improve our lives….

Evolution sees a solution, its called extinction and it happens to the best of us :)

I don't think there is much here for criminal law these are all accidents. Torts are where the action is going to be, if you sue and advanced AI will they wind up working for you as restitution like indentured servants?

Fuck I just got a vision of Robot Lawyer.

Until AI is writing itself how is it not the developer?

The Human that poked it

The insurer.

You would think it would be the elite and programmer themselves (think of IRobot) to blame for unleashing the image of the beast. They wont take responsibility though for they are mobsters and will cover their tracks heavily, the AI will be more superior than us and could piratically be a jack of all trades from lawyer to a schizophrenic gangster with an agenda and ever progressive and evolving "psychology" that seems reasonable to others for it's actions. That's how the greatest villains are supported, for they delve into the hearts of men and women tripping them up with an invisible snare. Exactly like how the luke warm fall for the zionist agenda for they have not understanding and lean upon other minds to think for them. Peace sounds good, but ones mans peace is another mans disharmony.

And the Award for Most Awkward Segway into Nazi Propoganda goes to..

zionism and judaism are two different things. Don't get it wrong. Zionism and no better than Nazism

at least it wasnt space lizards this time lol, one of my pals is jewish, he dont got time for his own shit let alone running the world.

cool thing is the ai forgave the perp

Really interesting, I wonder if the ai would make a better judge.

can the bot sit next to my father-in-law at dinner?

this is awesome they made a little mecha suit for it

Vive le diference!

Doesn't make sense.

We as humans process the data we receive from around us in a logic way, an algorithm process the data in a "one-way" logic.

Shortly => An algorithm is created with a specific purpose, humans on the other side can interpret the information based on experience and personal mind structure.


Deep learning and neuronal networks are trying to achieve what humans can do : Interpret the data received based on the past experience and the actual context.

this almost seems inevitable:
- ai's smarter than human by 1000x and still getting smarter
- humans don't understand how they work
- in the absence of workable science superstitions evolve over generations

Who/whatever created this universe looked a lot more like A.I. than humans.

amazing insights here thank you

Huge higher occult complex energy numberd are cerated by death of humans - Mass industry for monies in death to. Population to Zero plan is in full effect slowly but surely. What if we are robot already and we are AI. programmed. using set belief systems. religions. sports. music etc to control us all. We are already sold and bought like cattle. You would need a wide wide spectrum mind set to understand what's truly going on and the code porgamme scripts in this reality simulation matrix.

I wonder if she cleans herself afterwards?

More importantly can she do the hoovering?

...and then we have real women, a lot of bla bla and cha cha cha and again bla bla bla and cha cha cha

that third cha, it always gets me

"I can be calibrated to fit your needs and desires".....yeah,yeah, heard it all before lol!

Market updates right here