Anthropic has launched Claude 2.1, its latest and most advanced language model. By far the biggest and most noteworthy difference between Claude 2, which launched earlier this year, and Claude 2.1 is the amount of information that can be relayed to Claude via its message box.
Previously 128,000 tokens, Claude’s context window has now been enlarged to 200,000 tokens.
That’s around 150,000 words, or 500 pages of material – roughly equivalent to one cover-to-cover copy of Charlotte Bronte’s Jane Eyre.
Anthropic says that Claude 2.1 will be able to “summarize, perform Q&A, forecast trends, compare and contrast multiple documents, and much more” and expects the latency to “decrease substantially” as the technology develops.
In the wake of the generative AI (GenAI) revolution, UK businesses find themselves at a crossroads between unprecedented opportunities and inherent challenges.
Paul O’Sullivan, Senior Vice President of Solution Engineering (UKI) at Salesforce, sheds light on the complexities of this transformative landscape, urging businesses to tread cautiously while embracing the potential of artificial intelligence.
Generative AI has stormed the scene with remarkable speed. ChatGPT, for example, amassed 100 million users in a mere two months.
“If you put that into context, it took 10 years to reach 100 million users on Netflix,” says O’Sullivan.
This rapid adoption signals a seismic shift, promising substantial economic growth. O’Sullivan estimates that generative AI has the potential to contribute a staggering £3.5 trillion ($4.4 trillion) to the global economy.
“Again, if you put that into context, that’s about as much tax as the entire US takes in,” adds O’Sullivan.
One of its key advantages lies in driving automation, with the prospect of automating up to 40 percent of the average workday—leading to significant productivity gains for businesses.
One of the most obvious outcomes from the covid pandemic & the turbulent political environment over past few years is a recognition that we all tend to cling to our opinions and beliefs very, very tightly.
“I just know it” is often said; sometimes in the face of overwhelming evidence to the contrary.
But where does this certainty come from and just how sturdy is that foundation?
There is a word for this phenomena.
The Dunning-Kruger effect is a cognitive bias where people with limited competence in a specific domain overestimate their abilities. It occurs because they lack self-awareness and cannot accurately assess their own skills. Essentially, they think they know more than they do.
The initial study by David Dunning and Justin Kruger in 1999 focused on logical reasoning, grammar, and social skills. It revealed how people often misjudge their abilities due to a lack of metacognition (the ability to analyze one’s own thoughts).
Fact is, our thoughts and opinions are actually related closely to our memory.
Typically our opinions, our beliefs and our sense of self comes from,
Cognitive biases: These relates to systematic errors in thinking that affect how people process and interpret information; and they certainly influence their beliefs and decisions. Many cognitive biases are related to memory, such as confirmation bias, which is the tendency to seek out and remember information that confirms one’s existing opinions, or hindsight bias, which is the tendency to remember past events as more predictable or consistent with one’s current views than they actually were.
Emotional memory: This is the memory for events that are emotionally arousing or significant. Emotional memory is enhanced by the release of stress hormones that act on the amygdala, a brain region involved in processing emotions and fear. Emotional memory can affect opinions by making them more vivid, persistent, and more influential on behavior. However, emotional memory can also be inaccurate or intrusive, leading to false or distorted recollections of events that may affect opinions.
Narrative memory: This is memory for autobiographical events that are organized into coherent stories that give meaning and context to one’s life. Narrative memory is influenced by one’s concept of self, goals, values, and culture. Narrative memory can affect opinions greatly by shaping one’s identity, worldview, and sense of purpose. However, narrative memory can also be selective, reconstructive, and influenced by external sources, leading to inconsistencies or gaps in one’s personal history that may affect opinions.
So, if our opinions are linked inexorably to our memories. How good is yours?
Perceiving oneself as having a great memory is critical to every individual’s confidence, well-being, and daily functioning.
When we trust our memory, we approach tasks with confidence, whether it’s recalling information needed to do your job, approaching an exam or remembering important dates.
Feeling that our memory is reliable boosts our overall self-esteem and encourages us to take on challenges.
Memory enhances social interactions. Remembering names, details, and shared experiences makes us more likable and relatable.
Workplace: Professionals rely on memory for productivity. From meetings to deadlines, memory competence contributes to success.
Reducing anxiety: Fear of memory lapses can cause stress. Feeling confident in our memory alleviates anxiety.
Nostalgia: A vivid memory lets us relive cherished moments, fostering positive emotions.
And, a good memory accelerates learning. We build on past knowledge, adapt, and innovate.
This all leads us to want to over-estimate just how great our memory is.
I know that we all want to feel that our own memory is an amazing recording device; accurately taking note of the details of everything that we have experienced, read, visited and heard about. But the unfortunate fact is that it is not!
Elizabeth Loftus, one of the World’s leading experts on human memory, says, “If memory is the bedrock of the self … it is built on shifting sands.”
No-one has done more than psychologist Loftus (from the University of California, Irvine,) to expose the fallibility of human memory. In the 1990s, amid growing panic over claims of satanic abuse rings, the psychologist showed how easy it is for people to develop false memories of events that never ever happened.
All it took was being asked repeatedly to imagine them. At the time, unfortunately, this was a common psychotherapy technique to recover supposedly ‘repressed memories’.
Over the past three decades, Loftus, has become well-known for her work as an expert witness in legal cases. Her ongoing research on the fallibility of eyewitness testimony has taken on fresh importance in an era of fake news, the Me-Too movement and digital image manipulation. (1)
Elizabeth Loftus has found ordinary memory to be so malleable that she can easily prompt volunteers to “remember nonexistent broken glass and tape recorders; to think of a clean-shaven man as having a mustache, of straight hair as curly, of hammers as screwdrivers, to place a barn in a bucolic scene that contained no buildings at all, to believe in characters who never existed and events that never happened.”
“We now have a new paradigm of memory,” notes Loftus, “where memories are understood as a creative blending of fact and fiction, where images are alchemized by experience and emotion into memories.” (2)
So, we have established that our opinions, the facts we just KNOW, the statements we make, the feelings we believe, in fact most of our day-to-day activity relies heavily on memory.
So how chilling is it that this memory of ours is so malleable, that it changes over time, that it can be easily re-made by suggestion, and is so tenuous.
I guess you could say that a lot of the things we truly ‘believe’ could well be hallucinations of sorts?
Much like AI’s sometimes do.
AI ‘hallucination’ refers to a phenomenon where large language models (LLMs), such as OpenAI’s GPT-4, Bard AI or Google’s Palm, can generate outputs that contain false or misleading information presented as fact.
These outputs are sometimes nonsensical or altogether inaccurate and may not align with real data or events at all. Sounds human to me.
These are often reported on and discussed as a major hurdle to the use of artificial intelligence – but just how different is this really to asking a business colleague for help and assuming they actually know what they are talking about?
Or perhaps relying on your own memory to solve a problem assuming that your memory must obviously be correct?
If human memory and knowledge is as fallible as the ‘facts’ suggest, I feel that we may be giving AIs a bad rap! It may be that they are far more human than we would like to admit!
Amazon Bedrock goes GA with new AI models in the mix, plus AI-assisted coding customized for your organization, a generative AI boost for business analysts, and more (free) ways to get your AI skills dialed.
Amazon Web Services (AWS) announced five generative artificial intelligence (generative AI) innovations, so organizations of all sizes can build new generative AI applications, enhance employee productivity, and transform businesses.
“Over the last year, the proliferation of data, access to scalable compute, and advancements in machine learning have led to a surge of interest in generative AI, sparking new ideas that could transform entire industries and reimagine how work gets done,” said Swami Sivasubramanian, vice president of Data and AI at AWS. “Today’s announcement is a major milestone that puts generative AI at the fingertips of every business, from startups to enterprises, and every employee, from developers to data analysts. With powerful, new innovations AWS is bringing greater security, choice, and performance to customers, while also helping them to tightly align their data strategy across their organization, so they can make the most of the transformative potential of generative AI.”
Ensuring that AI does your bidding means getting the Prompt right. This Cheat sheet from AIFire.co provides a great guide to follow when building your prompts!
In a surprising turn of events, OpenAI, the renowned organization dedicated to developing artificial intelligence responsibly, finds itself at the center of controversy. The founder’s recent departure, allegedly due to a lack of transparency in internal communications, has raised eyebrows and sparked discussions about the intersection of ethics and profit in the realm of artificial intelligence.
The Departure:
Over the weekend, OpenAI’s founder stepped down amid accusations of inconsistent candor in communications. The vague reasoning behind this decision has fueled speculation on social media, leaving both the tech community and the general public curious about what might be happening behind closed doors.
The Stakes:
The significance of this shakeup extends beyond the internal dynamics of OpenAI. Artificial intelligence, often considered a potential existential threat to humanity, demands trustworthiness and stability from those leading its development. OpenAI, originally positioned as a not-for-profit organization, faces scrutiny as questions arise about the compatibility of responsible AI development and profit-driven motives.
Altruism Meets Capitalism:
OpenAI’s mission to prioritize responsible tech development over profit is now in the spotlight. As the world’s media closely watches the aftermath of this leadership change, concerns about potential hidden agendas within the organization may come to light. If the conflict indeed revolves around prioritizing profit over ethical considerations, it raises broader questions about the feasibility of balancing responsible AI with profit-driven initiatives.
Microsoft’s Role:
With Microsoft owning a significant 49% stake in OpenAI, the tech giant is now under increased scrutiny. Not only did Microsoft absorb two ex-OpenAI entrepreneurs, but it also granted the ousted founder a position within its ranks. As the drama unfolds, Microsoft’s potential influence over OpenAI’s future is a topic of speculation. The private company, valued at around $90 billion, may face further upheaval as employees express discontent over recent changes.
The clash at OpenAI serves as a stark reminder of the challenges inherent in navigating the delicate balance between ethics and profitability in the development of artificial intelligence. As the tech community awaits further developments, the incident prompts broader reflections on the future of responsible AI and the role of profit-driven interests in shaping the trajectory of this transformative technology.
The Elements of AI is a series of free online courses created by MinnaLearn and the University of Helsinki. We want to encourage as broad a group of people as possible to learn what AI is, what can (and can’t) be done with AI, and how to start creating AI methods. The courses combine theory with practical exercises and can be completed at your own pace.
