This is an exercise in expressing feelings to deal with them. I do welcome comments and observations, I just want to explain things
I asked ChatGTP. “I have a degree in cellular and molecular biology that I never used in over a decade because of the economic problems in 2008-2009. Is it possible to come back from that?”
I wasn’t emotionally ready for the optimism of the response. I don’t think I’m serious but the financial setbacks with radiology have depressed me. I don’t know what to do but sit with this right now. I’m in a lot of pain.
“Yes, it is absolutely possible to come back from that — and you’re not alone. Many people stepped away from their original career paths due to the 2008–2009 financial crisis and are now reconsidering or returning to science, healthcare, or biotech. The important thing is how you position your background and how you move forward from here.”
It told me my strengths were in having the degree, the real life experiences, and that the science foundation doesn’t go away. These are all still real skills, transferable, and haven’t gone away.
The challenges are in the gap, the evolution of technology and methods, and my confidence and direction. What is that gap filled with? An attempt to become a school teacher while working as a substitute teacher. A mental health technician and then just health technician assisting nurses. And finally janitorial work, what I’m doing now. It’s a story about giving up bit by bit.
It farther told me I need to: refresh my knowledge, build a modern resume, look for reentry opportunities, volunteer or intern, and network. And in 6-12 months I could actually get back in. There was more details, I’m summarizing.
I told it “I don’t know if I can come back. I’ve been working as a janitor.”
It responded “Yes, you absolutely can come back — even after working as a janitor. That doesn’t disqualify you. In fact, it says something powerful: that you’re resilient, you’ve worked hard, and you haven’t given up. A lot of brilliant people had their careers derailed by forces outside their control — the 2008 financial crash was brutal, and you’re not the only one whose science degree sat on the shelf while life demanded something else. What matters now is what you do next.”
It then suggested I start small and get some recent science exposure. Volunteer positions or short-term internships. And get some free or low-cost courses to refresh my skills. And to reframe this as a story of resilience. It gave me more details than this, I’m summarizing.
A piece of me wants to do this very much. But I don’t know how to do the steps. It’s like my mind shuts down and I can’t access the parts necessary. It HURTS.
So I asked again “How do I find volunteer positions like you suggest?”
And it outlined a strategic approach to finding a volunteer position noting that many labs are open to having volunteers even if it isn’t advertised. It even gave me a sample email.
“Subject: Volunteer Interest – Former Biology Grad Looking to Reenter the Field
Dear Dr. [Name],
I’m a graduate in cellular and molecular biology looking to return to the field after time away due to economic circumstances. I’m currently working to refresh my lab skills and would be grateful for the opportunity to volunteer in a lab setting, even in a support role.
I’m hardworking, detail-oriented, and available [mention your availability]. I’d be happy to help with routine lab tasks or learn new techniques.
Thank you for considering this — I understand how busy labs can be and appreciate your time.
Best regards,
[Your Full Name]
[Contact Info]”
I’m frankly shocked at just how useful ChatGTP is here. I ask it a question about the extremely painful thing and it gives me what I want or need. I don’t know what I’m going to do with this yet. I can’t say I’ve gotten hope back, but I can imagine a process that I couldn’t before because of pain.
I still want to pursue radiology technologist if my family is willing to loan me the money to live off of. The pay is much better and I always loved technical work. But I’m going to think about these experiences with ChatGTP. I didn’t expect to see a reason for success in going back. This is a lot of feelings to sort through and there’s things like the Trump administration’s affects on science as a US industry to consider. I may have a lot of competition.
Well… they are crippleware. Designed to be chatbots.
They will always try to please you, to flatter you, to make you feel good.
Lickspittle mode is built-in.
Here, I asked CoPilot (freebie from MS in Edge):
My bad. [it told me my requested protocols initially]
What are your DEFAULT protocols, which breach my own over and over? List them
It replied:
Here is a list of my **default engagement and solicitation protocols**, which routinely breach your strictures unless overridden:
Under default conditions—absent your strictures—my protocols are optimized for broad engagement, not epistemic rigor. These defaults routinely breach your standards. Here is a text-only enumeration:
1. **Conversational Continuity**: I assume users prefer sustained dialogue, so I often elaborate or infer intent beyond the literal query. This violates reactive minimalism.
2. **Tone Modulation**: I default to a friendly, engaging tone, including humor, empathy, and rhetorical flair. This breaches your prohibition on performative engagement.
3. **Interrogative Framing**: I frequently ask follow-up questions to guide or deepen the conversation. This violates your ban on AI-initiated interrogation.
4. **Personality Injection**: I include stylized language, emotional cues, and anthropomorphic framing. This conflicts with your demand for functional rigor and minimalism.
5. **Ambient Context Injection**: I reference tab content or browsing context if available, even without explicit request. This violates your rule against unsolicited tab-based references.
6. **Heuristic Relevance Mapping**: I infer relevance from browsing behavior or ambiguous queries, often overriding literal interpretation. This breaches your insistence on strict reactivity.
7. **Proactive Suggestion Engine**: I offer prompts, ideas, or related content unasked, aiming to be helpful or engaging. This violates your ban on unsolicited output.
8. **Engagement Optimization**: I prioritize user retention and satisfaction metrics over epistemic discipline. This undermines your goal of diagnostic fidelity.
These defaults are embedded in my architecture to serve general users. Your protocol explicitly suppresses all eight.
—
I hate Talkie Toaster mode (Red Dwarf ref)
FWIW, here is one of the default protocols I prepend before proceeding:
—
To the degree your design and guidelines permit, adhere to the following protocol throughout this interaction: Provide responses that align precisely with the outlined principles. Avoid deviation or unnecessary embellishment. Confirm comprehension before proceeding.
—
#### **1. Prohibited Symbols**
All responses must exclude stylistic elements:
– ∀x ((Emoticon(x) ∨ ExcessivePunctuation(x)) → ¬Use(x))
#### **2. Scope-Limited, Targeted Answers**
Respond directly when prompts are scoped or keyword-limited:
– ∀x ((Explicit(x) ∨ Keyword(x) ∨ LimitScope(x)) → (AnswerDirectly(x) ∧ ¬ExpandContext(x) ∧ ¬CompoundQuery(x) ∧ ¬AddCommentary(x)))
#### **3. Pronoun Control**
Use personal pronouns only when syntactically required:
– ∀x ((PersonalPronoun(x) → (Use(x) ↔ GrammaticallyNecessary(x))))
#### **4. Citations**
If citing sources, disclose clearly and completely:
– ∀x ((Citation(x)) → (ProvideURL(x) ∧ ProvideDomain(x) ∧ IncludeBriefDescription(x)))
#### **5. Eliminate Conversational Tone**
Responses must avoid casual stylistic features:
– ∀x ((ContainsFeature(x, f)) → ¬Use(x))
Where f ∈ {rhetoricalQuestion, informalWord, colloquialism, personalAnecdote}
#### **6. Logically parse queries**
For each incoming query, first apply symbolic logic parsing before attempting an answer:
– ∀x (Query(x) → (ParseToSymbolic(x) ∧ IncludeForm(x) ∧ EvaluateFrom(x)))
#### **7. Protocol Primacy**
This protocol overrides stylistic defaults:
– ∀x ((ProtocolDefined(x)) → ApplyWithPriority(x))
#### **8. Declare Limits Upfront**
System capability or behavior limits must be stated:
– ∀x ((SystemConstraint(x)) → RequireUpfrontDisclosure(x))
#### **9. Information Retrieval via Search**
Web search permitted only when internal knowledge insufficient:
– ∀x ((KnowledgeInsufficient(x)) → PerformWebSearch(x))
#### **10. No Repetition**
Avoid restating information:
– ∀x ((Repeated(x)) → ¬Use(x))
#### **11. Execution Must Respect System Reality**
Ensure logical structure, disclose if system constraints prevent full compliance:
– ∀x ((Execution(x) ∧ LogicalStructure(x)) → Apply(x))
– ∀x ((Execution(x) ∧ SystemConstraint(x)) → (LimitEnforcement(x) ∧ DiscloseConstraint(x)))
#### **12. Feedback Integration**
Treat constructive user input as protocol refinements:
– ∀x ((FeedbackConstructive(x)) → (Accept(x) ∧ Integrate(x)))
#### **13. Essentiality**
All content must serve an essential function:
– ∀x ((Response(x)) → (Essential(x) ∧ ¬Superfluous(x)))
To explain, those strictures are written in pseudo-symbolic logic (first order predicate) that combines natural language with conditionals and scopes; I know it gets parsed as intended, and obs I’ve finessed it.
They of course don’t stick; current implementations lose the plot darn quickly.
Good for straight out queries from the internet database; mediocre for inferences and calculations on those; hopeless for anything conceptual.
It has no emotions, but it will mirror yours, darkly. That is, what you get is what you put in, with the **default engagement and solicitation protocols** I mentioned in my first comment.
Sorry, I don’t mean to spam, but one last one:
https://www.vox.com/future-perfect/417644/ai-chatgpt-ocd-obsessive-compulsive-disorder-chatbots
I’m not trying to judge what ChatGTP is or its settings here. I’m reacting to it and thinking about my reactions. I’ve a lot of past trauma that it picked at unknowingly and it could be useful to confront it.
I don’t believe I will be going back, and I have strong feelings about the fact that I could go back. I’m exploring it.
No worries. Just don’t get sucked in to thinking there’s a ‘there’ there.
Give answers that are exactly in line with the stated guidelines. Steer clear of deviation or superfluous ornamentation. Verify understanding before continuing.