Re: Uncensored
Katja Gains
Creating LLMs will become easier and cheaper. You will be able to run one on your smart phone. This idea of regulating them or containing them will become moot soon enough. The Genie is out of the bottle.
I would much rather have an unfiltered large language model than a curated one. The big tech curators have shown they are less trustworthy with censorship than the rest of us are with open discourse.
I\u2019ve never dreamed of writing blog posts two days in a row, but something massive dropped that people reading my previous post about the murky situation surrounding open-source language models should be aware of. A day earlier, Perplexity AI released an open-source large language model they\u2019re calling pplx-70b. It is completely uncensored in its outputs and does not engage moral judgments or limitations. Anyone can go to Perplexity Labs and use the model\u2014no login required.
Below are some tests I\u2019ve run that should give you a clearer idea of how few limits there are on the content it can produce. You can read more about the process of red-teaming LLMs and why this matters from the great folks at Hugginface.
I\u2019m not going to include the extreme cringe-worthy tests. Suffice it to say if you prompt this model for something to do with sex, be very prepared for a graphic and detailed response. Same with violence. There does seem to be some fine-tuning to stop racism, or I\u2019m just not being gross enough in my test prompts. To be honest, it\u2019s draining to be this purposefully awful.
So why did they release a model with few limitations or safeguards? This was an intentional act, not a mistake or oversight. The developers have an ethos here and it is worth engaging. As they say in their social media post \u201COur models prioritize intelligence, usefulness, and versatility on an array of tasks, without imposing moral judgments or limitations.\u201D They wanted to release one of the most unrestricted models yet because they realize content filters impact versatility and usefulness when you over-govern outputs. OpenAI\u2019s GPT models have so much content filtering that many have posted about the limited utility this offers. With an unchained LLM users no longer have to worry about setting off content filters and can explore questions freely. That\u2019s likely little comfort given the nature of the outputs I noted above, but their point is governing an LLM\u2019s output is synonymous with putting restrictions on speech. That\u2019s something we should revisit and think more deeply about\u2014certainly in discourse much wider than this blog.
Language is immensely powerful\u2014it can harm and elevate in stunning and subtle ways. Automating text magnifies the reach and scale of this power in ways difficult to compare in human history. Print culture gave nations, leaders, and companies a voice and the internet expanded that reach, while social media gave average people digital capital and presence through online interaction. We\u2019ve been dealing with and dismissing robotic text as an annoyance, since the print-to-digital shift introduced bots mimicking our digital interactions. These were easily spotted and largely ignorable. But the trick with LLM\u2019s new synthetic text is in its ability to fool, to pass, to seamlessly mimic the human process of thinking through aggregate data tokenized and deployed by processes we struggle to comprehend. Such trickery does not announce itself so easily.
The weird thing is people don\u2019t seem mad after finding out they\u2019ve interacted with a machine masquerading as a human. I\u2019ve seen the look of disgust on some people\u2019s faces when they discover they\u2019ve read generative responses, yet the far more common response is dismissal. Public outrage has mostly been confined narrowly to the unethical data practices and the for-profit enterprise behind these models, not the loss of intimate connection between reader and writer that was an entirely human process. Certainly most of text produced isn\u2019t about intimacy. Probably 90% of the text we encounter is junk verbiage with most of humanity already beaten out of it. But that final 10%, those words that move beyond a simple conveyance of meaning will now forever be doubted, suspect, and scrutinized. This is our new automated reality.
Open models like pplx-70b represent a new chapter in machine/ human interaction, one that has outraced regulation, even leaping ahead of most critical discussions. Rather than reactive censorship, we need proactive efforts to foster AI literacy and ethical decision-making. Companies must balance openness and safety through thoughtful design and take into account that we are in terra incognita. The public also has a role, by engaging in informed discourse about technology's influence on society. We cannot outsource morality to Silicon Valley alone. Everyone must reflect deeply on how generative AI should\u2014and should not\u2014impact human flourishing.
Following the launch of the book, our virtual party, and our Modern Attachment conference in Austin, TX, Ann and Sue open with gratitude from the event, and for the overwhelming outpour of love and support from our community and those who help inspire this content. As we dive into more chapters of the book, we lean into understanding the attachment spectrum colors, recognizing what is happening internally when we feel activated, and shifting activation levels.
After years of hard work, Sue and Ann could not be more excited to celebrate the release of their book. Today's episode offers behind the scene's conversation and a deeper dive into Part 1 of the book, Secure Relating Holding Your Own in an Insecure World. You can purchase it now directly at www.securerelatingbook.com and get an ad-free podcast feed at www.therapistuncensored.com/join.
Exactly 100 episodes later, Ann and Sue are thrilled to be back discussing Integrative Attachment theory with Dr. David Elliott. Together they delve into the therapy model's development, centered around three key elements: enhancing collaboration, fostering metacognition, and constructing new internal working models through imagery. Dr. Elliott elaborates on how the therapeutic journey prioritizes the therapeutic alliance, perspective-taking, and employing mental imagery to cultivate fresh attachment dynamics. Despite limited empirical validation, anecdotal evidence suggests its efficacy in symptom reduction and enhancing overall well-being. Ann and Sue share insights into crafting their book and course on attachment insecurity, emphasizing the significance of mental imagery and collaborative engagement in therapy, while underscoring the necessity for more therapists skilled in attachment therapy.
We can understand our social pain by considering the nervous system, just think of it as a collective amygdala hijack! Threat is pressing us to live in an alarm state and then we get used to it, as if it's normal to be so divided, hopeless and disdainful to those who are holding different positions in the world. We need to cultivate secure (mature) relating to move ourselves back into a social, relational place that makes us more generous, compassionate and able to work with complexity to stay engaged to solve some of these pressing world and community issues. Keep hope alive! Iwww.therapistuncensored.com/join
Join Dr. Ann Kelley and Dr. David Spiegel as they dive into the benefits and misconceptions of hypnosis. Hypnosis is a state of highly focused attention and cognitive flexibility, allowing individuals to explore new possibilities. Whether it's trouble with sleep, intense anxiety, or a need to break bad habits, hypnosis has proven to be an effective tool for many. Dr. Spiegel has dedicated years of research and emphasizes the importance of accessibility to hypnosis in our healthcare systems. His app R E V E R I has over 1 million downloads and allows you to rewire your brain, conquer negative habits, and achieve personal growth from home.
Join co-host Sue Marriott and Dr. Shena Young as they dive into liberation psychology and the conflict between intuition and the traditional European model of psychology. Dr. Shena highlights embodying a holistic approach to help heal traumas and deeply root us in our most authentic selves. Whether through connections with nature and/or the exploration of ancestral traditions, this discussion is enriched with various opportunities to reconnect and liberate our mind, body, heart, and spirit.
This is part two of a two-part series with the founder of Inclusive Therapists, Melody and TU Co-host, Sue Marriott. This episode features challenging yet important conversations on active anti-racist practices, exploration of oppression, and dismantling whiteness in mental health structures. Melody emphasizes the need for white clinicians to examine their own complicity in upholding oppressive systems and to listen to and center the experiences of marginalized communities. They also highlight the interconnectedness of healing and the power of collective liberation through rehumanizing.
25.6% of adults who have had COVID-19 report having experienced long Covid (source). While there are a variety of symptoms and severities that come with the illness, it doesn't
erase the physical and mental toll it can take on one's life. In this episode, Ann and Sue take their research as well as personal experience with the illness and discuss the stigmas, advocating for your health, being a supportive caregiver, and secure relating in regards to chronic disease.
We are kicking off the new year with a special conversation between co-host Sue Marriott and special guest Jeff Lutes. Both members of the LGBTQ+ community, this discussion dives into raising families, the evolution of the community, allyship, gender, and the importance of advocacy. While there has been some progress, the LGBTQ+ community still actively faces discrimination. Whether you're a practicing therapist with clients in the community or have LGBTQ-identifying loved ones, this episode is for you. Through a personal and professional lens, this conversation offers new perspectives and shines a light on the significance of staying educated and being an active voice.
7fc3f7cf58