We have identified four topics that we believe are most significant and representative for AI in the year of 2022. These are heterogeneous, and \u2013 as may be expected for such a variable and complex field \u2013 they frequently overlap with each other. Taken in combination, however, they offer the most comprehensive picture of the state of the industry at the present time.<\/p>\n\n\n\n
Let\u2019s start with the first of these topics, by looking at a technology which you no doubt have already come across.<\/p>\n\n\n\n
1. Text-To-Image Takes Centre Stage, Points To Text-To-Video<\/h2>\n\n\n\n
The domains of Artificial Intelligence continued to expand this year, with some particularly surprising and exciting applications in mathematics<\/a> and the sciences<\/a>. But the field that saw the most memorable and globally impactful advances was that of Text-To-Image (TTI) generation<\/strong>, and this for several reasons.<\/p>\n\n\n\n One of these, inevitably, was its ubiquitous cultural fallout. Social media this year has been flooded with AI-generated pictures of all sorts, posted by users amused by the endless flexibility of these new tools and amazed at the results, while the case of a fine arts competition being won by an AI<\/a> has fuelled vibrant debates on the nature (and future) of art.<\/p>\n\n\n\n Another reason why TTI was so significant was the extent to which it was indicative of the broader trend of decentralisation in AI development (more on this below). While the most prominent TTI software this year belonged to Google (Imagen and Parti) and OpenAI (DALL-E 2), new AI Labs like Midjourney (which was used to win the aforementioned fine arts competition) and Stability.ai have also entered the fray and are doing surprisingly well<\/strong>, both in terms of technical achievements and economic performance. The race is not just to the swift this year, but to whoever wishes to run.<\/p>\n\n\n\n Finally, the TTI trend of 2022 points to the very likely next trend for 2023, which is that of Text-To-Video. Diffusion models this year definitively overtook Generative Adversarial Networks (GANs) for dynamic image generation<\/strong>, and as it turns out, these work quite well for video generation too. Meta\u2019s Make-A-Video and Google\u2019s Imagen Video (neither very imaginatively named) are both based on diffusion models, and they give us a taste of what to expect in the near future.<\/p>\n\n\n\n Commercial AI research has, for the last few years, been a domain almost exclusive to a few big players like OpenAI, Google\u2019s DeepMind, and Meta. However, 2022 may be remembered as the year when all of this changed, as two separate trends converged to suggest a rapid \u2013 and largely unpredicted \u2013 decentralisation of the industry.<\/p>\n\n\n\n Firstly, a great deal of the leading engineers working for the aforementioned \u2018big players\u2019 have left those companies and moved on to create their own AI startups<\/strong>. This includes, for example, almost all of the authors of the groundbreaking Attention Is All You Need<\/em> paper<\/a> (2017), formerly employed at Google and now heading smaller collectives like Adept<\/a>, Co:here<\/a>, Character.ai<\/a> and Inceptive<\/a>. (Meta did not see a similar exodus, but note that they already proactively decentralized their AI research<\/a>.)<\/p>\n\n\n\n On the whole, these new companies have been remarkably successful, particularly in the context of a year of global economic difficulties which saw investment in startups slowing down compared to 2021. For example, Anthrop\\c raised a staggering $580 million this year<\/a>, Inflection managed $225 million<\/a>, and Co:here $125 million<\/a>.<\/p>\n\n\n\n Secondly, open sourcing has successfully built on and refined on private AI models formerly believed to be beyond the technical skills of the collective<\/strong>. To mention just a few of various possible examples, this year we got Stable Diffusion<\/a>, Bloom<\/a> and OpenFold<\/a>, which all either replicate (or even improve on!) DALL-E, GPT-3 and AlphaFold 1, the models developed by OpenAI and DeepMind.<\/p>\n\n\n\n Stable Diffusion was developed by Stability.ai<\/a>, the entire business model of which is based on decentralisation and open sourcing. Their motto is \u2018AI by the people, for the people\u2019, and they are perhaps best described less as an AI lab than as a central hub for a variety of dispersed labs.<\/p>\n\n\n\n The most notable exception in this trend towards decentralisation is that of hardware. NVIDIA\u2019s chips still enjoy a stranglehold on the market, and there are no signs that this will change in 2023.<\/p>\n\n\n\n In June this year, Blake Lemoine, then a senior software engineer working for Google, made the sensational claim that the company\u2019s ultra-advanced chatbot LaMDA was self-aware<\/a>. Lemoine was subsequently fired, and his claims were dismissed by Google and by AI experts at large<\/a>.<\/p>\n\n\n\n The story made international headlines in a way that very seldom happens in this industry, in no small part thanks to the transcript published by Lemoine of his dialogues with LaMDA<\/a>. Here the engineer is seen querying the chatbot on the topic of consciousness, and the exchange is undeniably fascinating to read.<\/p>\n\n\n\n To dismiss this episode as a blunder with neither consequence nor scientific basis is to miss the point entirely. The real lesson for the AI industry here is the ease with which the indicators of consciousness can be replicated, or more accurately, falsified<\/strong> \u2013 to the point that a leading expert in the field can be so thoroughly bamboozled by a machine.<\/p>\n\n\n\n While we still seem to be relatively far off (emphasis on \u2018relatively\u2019) from developing self-aware Artificial Intelligence, the Lemoine controversy reminds us that we also have no reliable model to verify self-awareness \u2013 not even for animals, let alone machines. The question is not when will AI become sentient \u2013 but how will we know that it did?<\/strong><\/p>\n\n\n\n After an extensive period of ratification that began last year, the European Union is expected to finally pass the Artificial Intelligence Act<\/a> before the end of 2022. The law introduces mandatory risk assessment in AI development and regulates its commercial use.<\/p>\n\n\n\n The law itself is, in its current form, rather light \u2013 it is more a collection of general guidelines than anything else. But its significance for the international AI industry could be extremely broad. The EU\u2019s tech regulations have a tendency to be adopted globally by what is known as the Brussels effect<\/a>, as we\u2019ve already seen with the European Union\u2019s General Data Protection Regulation (GDPR).<\/p>\n\n\n\n This is not the only example of emergent and growing concerns \u2013 especially in Europe \u2013 related to AI safety. It follows on the National AI Strategy<\/a> document published last year by the UK government, which is notable among other things for its recognition of the existential risks<\/a> posed by superhuman artificial general intelligence (AGI), and the founding this year of Conjecture<\/a>, a British startup dedicated to mitigating those risks.<\/p>\n\n\n\n![\"\"](\"https:\/\/www.wbscodingschool.com\/files\/pic-1-better-c-Jason-Allen-1024x721.jpeg\")
2. The Industry Decentralizes (Unexpectedly)<\/h2>\n\n\n\n
![\"\"](\"https:\/\/www.wbscodingschool.com\/files\/2.-AdobeStock-by-Andrii-Yalanskyi_437288675-1024x614.jpeg\")
3. Sentient Google Chatbots \u2013 The Blake Lemoine Controversy<\/h2>\n\n\n\n
![\"\"](\"https:\/\/www.wbscodingschool.com\/files\/3.-AdobeStock-by-Tada-Images_510546305_Editorial_Use_Only-1024x682.jpeg\")
4. Regulation: A Tale Of Two Cities<\/h2>\n\n\n\n
![\"\"](\"https:\/\/www.wbscodingschool.com\/files\/4.-AdobeStock-by-tanaonte_181379251-1024x666.jpeg\")