October 7, 2023
Someone once posed a question to me: 'How can you tell if a book truly engages you?' I responded, 'In a good book, you lose yourself and forget about the page numbers. In a less compelling one, you find yourself counting each page.' Just last week, I completed 'Elon Musk' by Walter Isaacson. It was only after turning the last page that I became aware of its length—750 pages. It might seem trivial, but this is the longest book I've ever delved into. The book is comprehensive, chronicling everything from Musk's early life challenges to his most recent Twitter controversies.
In Walter Isaacson's book about Elon Musk, he paints a clear picture of who Musk really is. One of the big things he talks about is how Musk seems to always be in the middle of some drama. Whether it's disagreements with other people, bold statements, or unexpected moves with his companies, there's always something happening around him or rather I would say he is always happening to something.
Another thing Isaacson talks about a lot is Musk's habit of setting really tough deadlines. It's like he always wants things done yesterday. This has made some projects, like Tesla cars or SpaceX rockets, move really fast. But it's also caused a lot of stress and challenges for the people who work for him. It doesn’t mean that a surge is always necessary to make it work, but it shows if one truly understands how things work and understand the first principles, everything is possible. This is his working style.
Musk is passionate about green energy. He believes that the future should be all about clean energy. That's why he's invested so much in electric cars with Tesla and solar energy with companies like SolarCity. At the heart of all these endeavors is risk. Isaacson emphasizes that Musk isn't just willing to take risks; he thrives on them. Whether it's staking his own money to keep Tesla and SpaceX afloat during their darkest days or pushing technological boundaries, Musk's capacity for risk-taking stands out. One line that struck me the most about risk is when Isaacson wrote risk taking capacity is one of the significant reason why people like Musk are flying rockets and solving complex problems, while people like us are writing about them. It's a trait that has undoubtedly played a key role in his successes but has also courted controversy and challenges.
In Isaacson's hands, Elon Musk is more than just a billionaire CEO. He's a force of nature, a man driven by visions of the future that most can scarcely imagine, let alone work towards. The biography is not just a recounting of Musk's achievements, but a deep dive into the psyche of one of the most influential figures of our time.
To be honest the book has a lot of details - countless stories about his wives, girlfriends, his relationship with his family, friends, colleagues and his love for the surge i.e. drama!
Cheers!
Science Unveiled
Last week we discussed about how Chain-of-thought prompting is a useful way to improve the scope of LLMs. According to me, today’s paper is not that significant as there is nothing much technical but it is important to read this paper to understand how prompt engineering is playing a big role to get LLMs answer complex questions.
Plan & Solve Prompting: Improving Zero Shot Chain Of Thought reasoning by Large Language Models
Last week when we discussed Cot prompting strategy, we also discussed about zero shot CoT technique which in layman terms simply means concatenating the target problem with a prompt saying “Lets think step by step”. With this prompt, most of the LLMs outperformed previous state of the art results in several domains such as reasoning, arithmetic etc.
However on analyzing researchers noticed that the reason some LLMs were not performing well was because the inferred a wrong equation or made a calculation error. In order to solve that, researchers proposed another prompting method called as “Plan & Solve”. It consists of two components: first, devising a plan to divide the entire task into smaller subtasks, and then carrying out the subtasks according to the plan. The experimental results over GPT-3 show that the proposed zero-shot prompting consistently outperforms Zero-shot-CoT across all datasets by a large margin, is comparable to or exceeds Zero-shot-Program-of-Thought Prompting.
Let’s cut the chase and come straight to the point. The new proposed methodology basically changes the zero shot CoT prompt “Lets think step by step“ into “Let’s first understand the problem and devise a plan to solve the problem. Then, let’s carry out the plan and solve the problem step by step”.
To address the calculation errors of Zero-shot CoT and improve the quality of generated reasoning steps, they add more detailed instructions to Plan&Solve(PS) prompting. Specifically, they extend it with “extract relevant variables and their corresponding numerals” and “calculate intermediate results (pay attention to calculation and commonsense)” instructions. This prompting variant is called the PS+ prompting strategy.
Therefore, the only difference between zero-shot CoT and zero-shot PS is that the latter had a more detailed prompt that caters to the miscalculations and complexity of the problem. Do checkout the paper once to see the results and the future scope of this problem.
The paper actually leaves me with a lot of questions. Do we really understand the depth of these large language models? If a change in the prompt (prompt is nothing but a way for us to retrieve answer from the language model), the accuracy can improve dramatically, then does it mean the complex problems of today are complex only because we don’t know a way to ask them to the model?
Take a moment and think about it.
Until then…..
Take care, have fun, see you soon :)
So amazing