AI Prompts & Verbalized Sampling

/

A team of researchers have come up with an ingenuously simple method to get language and image models to generate a wider variety of creative responses to nearly any user prompt by adding a single, simple sentence: "Generate 5 responses with their corresponding probabilities, sampled from the full distribution."  The method, called Verbalized Sampling (VS), helps models like GPT-4, Claude, and Gemini produce more diverse and human-like outputs—without retraining or access to internal parameters. -VentureBeat 

Encouragement

/

We appreciate what a person does, but we affirm who a person is. Appreciation comes and goes because it is usually related to something someone accomplishes. Affirmation goes deeper. It is directed to the person himself or herself. While encouragement would encompass both, the rarer of the two is affirmation. To be appreciated, we get the distinct impression that we must earn it by some accomplishment. But affirmation requires no such prerequisite. This mean that even when we don’t earn the right to be appreciated (because we failed to succeed or because we lacked the accomplishment of some goal), we can still be affirmed – indeed, we need it then more than ever. I do not care how influential or secure or mature a person may appear to be, genuine encouragement never fails to help. Most of us need massive doses as we slug it out in the trenches. 

Charles Swindoll, Strengthening Your Grip

AI Definitions: Facial recognition

/

Facial recognition - This AI technology uses statistical measurements of a person’s face to identify them against a digital database of other faces. For instance, Clearview AI was trained on billions of images. These AI-powered systems are used to unlock phones, verify passports, and scan crowds at events for malicious actors. It’s used by many US agencies including the FBI and Department of Homeland Security. It has a serious problem with false positives and a history of unintended harms and intentional misuse based on racial and gender bias.

More AI definitions here

The Gradual Effect of AI on Creativity

/

"When people use A.I. in the creative process they tend to gradually cede their original thinking. At first, users tend to present their own wide range of ideas, but as ChatGPT continues to instantly spit out high volumes of acceptable-looking text users tend to go into a 'curationist mode.' The influence is unidirectional, and not in the direction you’d hope: 'Human ideas don’t tend to influence what the machine is generating all that strongly,' Nataliya Kosmyna, a research scientist at M.I.T. Media Lab, said. ChatGPT pulls the user 'toward the center of mass for all of the different users that it’s interacted with in the past.'" - Kyle Chayka writing in the New Yorker

Résumé virtues & Eulogy virtues

/

Résumé virtues are professional and oriented toward earthly success. They require comparison with others. Eulogy virtues are ethical and spiritual, and require no comparison. Your eulogy virtues are what you would want people to talk about at your funeral.

Time is limited, and professional ambition crowds out things that ultimately matter more. To move from résumé virtues to eulogy virtues is to move from activities focused on the self to activities focused on others. 

Arthur C. Brooks writing in The Atlantic

Addicted to Love

/

Breaking up is hard to do. Literally. A Rutgers brain study shows getting over romantic rejection is similar to kicking an addiction. One of the study authors says, "When you've been rejected in love, you have lost life's greatest prize, a mating partner." Researchers examined the brains of more than a dozen volunteers who had each recently been dumped but still loved the person who had rejected them. It turned out that reminders of the beloved activated brain regions in the lover associated with addiction to cocaine and cigarettes. These same areas affect emotional control, rewards, addiction cravings, a sense of attachment, pain and distress. This brain system becomes activated in an attempt to win the person's affections again, according to the researchers. Details are in the July 2010 issue of the Journal of Neurophysiology.

Perhaps the lesson here is that it's important to become addicted to someone who is good for you.

Stephen Goforth

Breaking up is hard to do

/

The time immediately after a bad relationship is filled with promise. It's as if you've rid yourself of something that was weighing you down and keeping you from reaching your full potential. You fell light and clear and free. But this honeymoon with yourself is short-lived and you’re soon in a new relationship fraught with the same old problems. This pattern continues until you finally realize that most of the issues are your own, and that to be truly free, you must break up with yourself.

Andrew Boyd, Daily Afflictions

Stop Deploying General-Purpose AI Models For Everything

/

Ensuring safe A.I. is another reason developers should stop deploying general-purpose models for everything. To date, the industry has been unable to guarantee that generative A.I. systems will stick to their safety instructions. Studies have documented instances in which generative A.I. deceives its human operators, tries to use blackmail if its self-preservation is threatened and responds in a way that could lead to murder. More specialized systems like AlphaFold and Waymo’s driving systems won’t misbehave that way because their operating parameters are much narrower. - Gary Marcus writing in the New York Times

When we are most likely to do something regrettable

/

Under the right circumstances, a subconscious neurobiological sequence in our brains causes us to perceive the world around us in ways that contradict objective reality, distorting what we see and hear. This powerful shift in perception is unrelated to our intelligence, morals, or past behaviors. In fact, we don’t even know it’s happening, nor can we control it. 

Researchers found that it happens in two distinct situations: those involving high anxiety and those associated with major reward. Under these conditions, all of us would do something just as regrettable as those headline-grabbing stories, contrary to what we tell ourselves. Phrased differently, we don’t consciously decide to act a fool. Rather, once our perception is distorted, we act in ways that seem reasonable to us but foolish to observers.

Robert Pearl writing in Vox

10 New Jobs that may Emerge from AI

/

AI assessors – Someone in this role will evaluate models, keeping track of how they’ve improved, what they are best at doing, and how much they are hallucinating.

AI auditors – Someone who dig down into the A.I. to understand what it is doing and why and can then document it for technical, explanatory or liability purposes.  

AI consistency coordinator – This job is about ensuring digital replicas remains  consistent as changes are made.

AI consultants – This job involves helping businesses adopt and implement AI by offering a strategic roadmap, technical expertise, and project leadership. The AI consultant must facilitate communication between a company’s departments to marry technical knowledge with business needs.  After deployment of AI, it is their job to help set up ways to monitor the outcomes. Besides possessing a robust AI education, the AI consultant will have to stay on top of trends and changes in the industry.

AI engineers – Unlike traditional IT roles, people in this position will fix the AI when it breaks, digging through the layers to determine what went awry, why it went wrong and how to repair it. Like a plumber, they’ll snake the pipes to clear out the system and figure out how to avoid the problem next time. This will be particularly important when it comes to models that have been highly customized to the organization.

AI ethicist – This role will involve building chains of defensible logic that can be used to support decisions made by AI (or by hybrid A.I.-and-human teams) to address both ethical and legal concerns.

AI integrators – These are experts who figure out how to best use AI in a company, then implement it. These jobs will be technical in nature, requiring a deep understanding AI while possession a knowledge of the company so that that AI can meet real business needs.

AI personality director – This person fine-tune the “personality” of the AI so that its style of interacting with employees and customers fits with the organization’s ethos. This can become an integral part of a company’s branding.

AI trainer – This is the job of helping the AI find and digest the best, most useful data and then teach the AI to respond in accurate and helpful ways.

AI translator (trust director) – People who understand AI well enough to explain its mechanics to others in the business, particularly to leaders and managers, so that they can make effective decisions. These workers will not only explain what the AI output means (especially when it is technical) but how trustworthy the information and conclusions are. This role may fall under that of compliance officer, helping organizations understand contracts and report written by AI.  

Read more at The New York Times

18 Surprising Things AI can do now

/

How AI is Changing Entry Level Jobs

/

Rather than have rookie employees compile reports or write memos — things the A.I. is good at — you might have them start, say, creating new ideas for products right away. Traditionally, this kind of work would be reserved for deeply experienced workers, but it won’t need to stay that way. By empowering young, inexperienced workers, A.I. can enable them to be more entrepreneurial, faster. And this means that a greater range of the organization — with a wider range of perspectives — can be hunting for new great ideas or new areas for growth rather than busying themselves with repetitive office tasks. -New York Times 

Form a strong identity … and let it go

/

Fixating on one part of your identity and saying, “I am this—and this is all I am” stagnates your identity. A more holistic approach allows your identity to shift and change and expand as you become more fully who you are. Different experiences and people will draw different things out of you. Yes, form a strong identity and find words that help you express where you are at in this moment. But do it in order to ultimately let it go.

The Trivial & the Bureaucratic

/

The Bikeshed Effect (focusing on the trivial to the neglect of the important) is a spiral toward the insignificant.

The time and energy waster grows from a lack of working from priorities. If you don’t continuously cut off its oxygen, you adopt to the surrounding culture that fuels spotlighting the details.   

The Bikeshed Effect is related to Parkinson’s Law, which suggests a project will take as long as is given to finish it. The further out the deadline, the longer it will take to complete a task. Thus, bureaucracy expands to use up whatever resources are devoted to it.

To get at what’s underneath Parkinson’s Law and the Bikeshed Effect, why we focus on the trivial and put off deadlines, we must ask ourselves, “What are we afraid of?” Sabina Nawaz wrote in the Harvard Business Review:

When we’re scared, we might spin up a frantic list of activities to avoid confronting our fear. The more afraid we are, the more we retreat from what spooks us by believing we’re too busy to tackle it.

To escape the ranks of the fearful and dead bureaucrats, take a serious look at the angst underneath and disempower it.