Phrasly vs SupWriter

Generative AI Will Change How We Do Knowledge Work Most people either believe generative AI will replace knowledge workers or think it will never reach that point. In reality it will transform how we work by making mundane cognitive work cheaper. Humans will become better at knowing what to ask for and making better decisions on what the output should be.

Generative AI has already proved that it is able to do many tasks. Writing emails, summarizing articles, coding basic functions and generating first drafts of anything you want written. People have spent countless hours doing these types of tasks throughout their careers. In one study they found that those using ChatGPT completed writing tasks in 40% less time and received higher scores on the output. Programmers also found themselves getting through their assignments quicker when using Github Co-Pilot. So what happens when we have access to instant first drafts?

Well first we need to figure out what problem we are trying to solve. We then need to verify if our AI gave us the right answer and know when its giving us false information. These models will make up fake references and can be extremely dumb when it comes to understanding your prompt. People have been fired when submitting papers with false references generated by AI. While these workers were not replaced by generative AI they failed to do their due diligence on verifying the facts they presented. As generative AI takes off we will all become responsible for making sure the output we give others is correct.

This new technology can be very dangerous to those starting out in a career. Many jobs rely on junior employees to complete basic tasks that will allow them to learn from the experience. We may have to develop different paths for learning if generative AI does all of these small tasks for us. Companies will continue to evolve on how they utilize this new form of AI but I believe we will see humans continue to do many tasks that require decision making.

There has been many advances made recently to help these systems better understand your question. There could be a point where these systems can answer your questions perfectly but I don't believe that's possible anytime soon.

# Generative AI Will Reshape Knowledge Work Not Replace It

Opinions are divided over generative artificial intelligence and knowledge workers. On one hand, are those who predict that millions of analysts, writers and programmers will be displaced by the technology. And on the other, the anti-noise. They dismiss this technology as a mere overhyped autocomplete. Both individuals overlook the more engaging reality. Repetitive knowledge work will be devalued by generative AI, while judgement, taste and accountability will become more valuable.

What are the abilities of the system? They write Emails, summarize documents, produce template code, and kickstart any text into a usable first draft. Once, they occupied large parts of a professional's day. A study by researchers at MIT has shown that workers using ChatGPT on writing tasks finished almost 40% faster with higher quality ratings. Developers using Copilot at GitHub complete their tasks faster than those without its assistance. The draft is no longer the valuable part of the job when it costs almost nothing to produce the first draft.

Everything surrounding the draft ultimately becomes valuable. Some must decide which of the problems are worth solving, check whether the output is right or wrong, and realize when a confident-sounding answer is wrong. Generative models create fake citations, misinterpret contexts, and flatter users. The lawyer who submitted a fake citation created by AI as has happened in the extremely well-known Mata v. Avianca case was not replaced by technology but outed for not exercising professional judgment. As generated content proliferates, verification and accountability remain human responsibilities that will become more important.

However, there is a real risk here – especially for newbies. Traditionally, entry-level knowledge work has served as a kind of apprenticeship: junior analysts learn by writing memos, and junior developers learn by fixing bugs. As AI takes over more tasks organizations will either need to find new ways of developing expertise or they will run short of seniors to supervise the machines. Organizations that utilize AI just for cost-cutting measures may rob themselves of their own talent pipelines.

Critics claim that capabilities are progressing so quickly after judgment-laden work will be automated too. We cannot exclude that possibility. But history shows that automation changes jobs, rather than destroying them, over time. Spreadsheets did not do away with accountants; rather, they saved them hours of arithmetic and broadened the range of things accountants were expected to analyse.

One cannot panic or be indifferent. Workers must develop skills in problem framing, critical evaluation, and domain expertise. Roles and training should be redesigned with human oversight in mind. Generative AI is the new powerful lever. Like all levers, the effects depend on who’s pulling and how carefully they understand what they are moving.