It’s no secret that AI & ML have been all the rage in recent years, finding applications in various business domains more frequently than ever. However, as the popularity of this technology soars, so do the associated risks. Here at LAB325, we’ve delved into the topic, conducting an in-depth study based on numerous publications and reports to unravel what lies ahead for AI & ML in the near future.

Artificial Intelligence, often referred to as AI, is a revolutionary process that involves constructing intelligent machines through vast amounts of data. These systems learn from previous experiences and knowledge to perform tasks akin to human capabilities. By augmenting speed, precision, and efficiency, AI complements human efforts. The foundation of AI lies in intricate algorithms and methodologies that empower machines to make independent decisions. At its core, AI is powered by Machine Learning and Deep Learning, paving the way for incredible advancements in technology. As the learning process progresses, the performance of ML algorithms dynamically enhances with a larger pool of available samples. Take deep learning, for instance, a sub-domain of machine learning that empowers computers to imitate human traits like learning from examples. It outperforms conventional ML algorithms, delivering superior performance parameters.

Artificial Intelligence can be categorized based on its capabilities and functionalities.

In terms of capabilities, there are three types of AI:
1. Narrow AI
Narrow AI, also known as Weak AI, is designed to handle specific tasks within its limited scope and cannot extend beyond those boundaries. It focuses on excelling in a particular subset of cognitive abilities, and its prevalence is on the rise as machine learning and deep learning techniques progress.
Examples of Narrow AI are widespread in our daily lives. Take Apple Siri, for instance, which operates with a predefined set of functions and may encounter challenges when faced with tasks outside its capabilities. Another notable Narrow AI is the IBM Watson supercomputer, which leverages cognitive computing, machine learning, and natural language processing to process information and provide answers. Notably, IBM Watson showcased its capabilities by defeating human contestant Ken Jennings on the renowned game show, Jeopardy!
Additional examples of Narrow AI encompass Google Translate, image recognition software, recommendation systems, spam filtering, and Google’s page-ranking algorithm. These applications demonstrate the practicality and effectiveness of Narrow AI in various domains.

2. General AI
General AI, or strong AI, possesses the ability to comprehend and learn any intellectual task akin to a human being. It empowers machines to apply knowledge and skills across various contexts. However, as of now, AI researchers have not succeeded in creating strong AI. The challenge lies in programming machines with full cognitive abilities, including consciousness, a feat yet to be achieved. The pursuit of General AI continues to drive innovation and exploration in the field of artificial intelligence.
Fujitsu has engineered the K computer, one of the world’s fastest supercomputers, making significant strides toward achieving strong AI. However, simulating a single second of neural activity took nearly 40 minutes, making the timeline for achieving strong AI uncertain. In the realm of supercomputers, Tianhe-2, developed by China’s National University of Defense Technology, holds an impressive record of 33.86 petaflops (quadrillions of cps). While this achievement is remarkable, it pales in comparison to the human brain’s estimated capacity of one exaflop, equivalent to a billion cps. The quest for AI’s advancement continues to push the boundaries of computational capabilities.

3. Super AI
Super AI, the pinnacle of artificial intelligence, goes beyond human capabilities and excels in every task. This concept of artificial superintelligence envisions AI reaching a level where it not only comprehends human emotions and experiences but also possesses emotions, needs, beliefs, and desires of its own. Currently, it remains a hypothetical existence. Key traits of super AI encompass independent thinking, problem-solving, and the ability to make judgments and decisions autonomously. As we explore the realm of AI, the potential for super AI challenges our understanding of intelligence and consciousness.

The various types of Artificial Intelligence are categorized based on functionalities.

1. Reactive Machine
Reactive machines form the fundamental type of artificial intelligence, relying solely on current data without storing memories or using past experiences for future actions. They perceive and respond to the immediate world around them. Reactive machines are designed for specific tasks and lack capabilities beyond those assigned to them.
IBM’s Deep Blue, which famously defeated chess grandmaster Garry Kasparov, exemplifies a reactive machine. Deep Blue can analyze the chessboard’s current state, identify pieces and their possible moves, and make predictions about future moves. However, it doesn’t refer to past experiences or improve with practice, focusing solely on the present moment to make decisions.

2. Limited Memory
Limited Memory AI, also known as Memory-Based AI, utilizes past data to inform its decision-making process. However, the memory of these systems is temporary, and they can only access this past data for a limited duration. Unlike humans who can accumulate knowledge over time, Limited Memory AI cannot retain and add past experiences to its repository. This technology finds applications in self-driving vehicles, where the system relies on recent data to navigate and make real-time driving decisions.

3. Theory of Mind
Theory of Mind AI is an advanced and theoretical form of technology that is currently conceptual. This type of AI would require a deep comprehension that individuals and objects in an environment can influence emotions and actions. It would need to understand human emotions, feelings, and thoughts. While significant advancements have been made in this area, Theory of Mind AI is still a work in progress and has not been fully realized yet.
A practical illustration of Theory of Mind AI is demonstrated through Kismet, a robot head created by a researcher from the Massachusetts Institute of Technology in the late 90s. Kismet can mimic and recognize human emotions, showcasing significant advancements in this field. However, it lacks the ability to track gazes or convey attention to humans. Another notable example of Theory of Mind AI is Sophia from Hanson Robotics. Equipped with cameras in her eyes and advanced computer algorithms, Sophia can see, maintain eye contact, recognize individuals, and track faces. Her capabilities highlight the progress made in implementing Theory of Mind AI.

4. Self Awareness
Self-awareness AI remains a theoretical concept for now. These hypothetical systems possess the ability to comprehend their internal traits, states, and conditions, as well as perceive human emotions. If realized, these machines would surpass human intelligence. They would not only understand and evoke emotions in their interactions but also possess emotions, needs, and beliefs of their own. This level of AI represents a fascinating and thought-provoking possibility for the future.

Diverse Branches of Artificial Intelligence
AI research has made remarkable strides, unveiling effective techniques to solve a wide array of problems, spanning from game playing to medical diagnosis. Within the realm of AI, numerous branches have emerged, each with its own focus and set of techniques. Here are some of the pivotal branches of AI:

Machine Learning: This field revolves around developing algorithms with the capacity to learn from data. Machine learning finds applications in image recognition, spam filtering, and natural language processing, among others.

Deep Learning: A subset of machine learning, deep learning employs artificial neural networks to acquire knowledge from vast datasets. It has proven immensely successful in solving challenges such as NLP, image recognition, and speech recognition.

Natural Language Processing (NLP): NLP centers on the interaction between computers and human language. It empowers systems to comprehend and process human speech, driving applications like machine translation, speech recognition, and text analysis.

Robotics: Robotics, a field of engineering, delves into the design, construction, and operation of robots. These automated machines undertake tasks in diverse industries, including manufacturing, healthcare, and transportation.

Expert Systems: Designed to emulate the reasoning and decision-making abilities of human experts, expert systems find applications in various fields, from medical diagnosis and financial planning to customer service.

Each branch of AI holds tremendous promise, propelling us toward a future brimming with innovation and possibilities.

Here is the list of the most popular AI platforms and services:

Text Applications

Search: Internet

Andi, Friday Go, Hotbot, Komo, Metaphor, Microsoft new Bing, Neeva

Search: Enterprise, Sales, Marketing & Accounting

ExactBuyer, Glean, Hebbia, keywrds.ai, Pinecone, Vectara, weaviate.io

Search: Jobs

Apply AI, hireBrain, JobHuntMode, MatchThatRoleAI, NetworkAI

Search: Books, Images, Podcasts, Videos, & TV

All Search, Anypod, BOOKCLUB.AI, Context, EveryPixel, Sona, Tapesearch, Twelve Labs, Watch Now

Search: Research

AlphaResearch, Elicit, NewsDeck, OpenRead, Semantic Scholar, SciSpace, WolframAlpha

Search: Programming/Software Development

Bloop, Phind, Context, GitHub Data Explorer, Buildt

Chat

OpenAI ChatGPT, Google Bard, Claude

Sales & Marketing Copy Generation

Anyword, copy.ai, copymatic, Creatext, Frase, Hypotenuse, Jasper, Lately, Letterdrop, Longshot, Peppertype, Persado, Podium, Regie, Rytr, SmartWriter, Writesonic

Email Generation

Autobound, Benchmark, HyperType, leadIQ Scribe, MailMentor, mails.ai, Rytr, Salesboom.ai

Other Copy Generation

Compose, copysmith, DeepReview, elaborate, HyperWrite, Jenni, Linguix, Scribe, Writer

Note-Taking & Document Summarization

A121 Labs WordTune Read, Cogram, genei, Hila, Notably, Scholarcy

Writing Assistance & Translation

A121 Labs WordTune, git18n, Grammarly

Image Applications

2D Image Generation

AI Images MagicStock, Artbreeder, Bing Image Creator, Craiyon, DALL-E 2, Midjourney, NightCafe, Shutterstock AI Image Generator, Solace Vision, Spellbrush, Stability DreamStudio, starryai, Stock AI, stockimg.ai, WOMBO Dream, Zazow

Web Design, Color Palette Generation

Colormind, Designs.ai, Fronty, Khroma, Mutiny, Uizard

Image Editing, Enhancement & Style Transfer

Autoenhance, Facet, Imagen, Let’s Enhance, PhotoRoom, Rosebud, Stability ClipDrop

Ad Generation

Blend, Collider AI, Memorable, Pencil, Re:cast

Presentations/Slide Generation

Autoslide, beautiful.ai, decktopus, MagicSlides, motionit, PowerMode, Powerpresent AI, presentations.ai, Prezo, SlidesAI

3D Image Generation

Bifrost, Common Sense Machines, Hypar, Kaedim, Maket, Sloyd

Digital People Generation

Botika, Soul Machines, Replika , anima, Colossyan, Metaphysic

Video Applications

Video Generation From Text

Runway, HeyGen, Synthesia, Pictory, Elai, Steve AI

Video Editing

PEECH, Pinterest VOCHI, Recut, VEED, Lumen5, Flexclip

Video Personalization & Derivative Content Generation

Exemplary.ai, Maverick, Tavus, Terra , Video Tap, vidyo.ai, Yepic

Audio Applications

Music Generation

Aimi, Aiva, AudioShake, beatoven.ai, Endel, HookGen, Mubert, Musico, OpenAI Jukebox, Soundful, Soundraw, Staccato

Text to Speech

Au Auris, Coqui, DeepZen, ElevenLabs Prime Voice, Listnr, Lovo, Murf, Replica, Resemble.ai, Speechify, Voicemaker, voicera, WellSaid

Speech to Text (Transcription)

Transcription: General

Deepgram, ermine.ai, MacWhisper, Revoldiv, Rhythmex, ScribeBuddy, Transkribieren, Type Studio, vid2txt, Vocapia

Transcription: Note Taking

Berrycast, Fathom, Good Tape, Read, Sembly , tl;dv, Zoom IQ

Transcription: Subtitle Generation

FREESUBTITLES.AI, izwe.ai, ScriptMe, Taption, Targum, Zubtitle

Transcription: Podcasts

Descript, Riverside, Steno

Transcription: APIs

AssemblyAI, Neural Space

Transcription: Other

EchoFox, Heark, krekicki, TranscribeMe, Video2Recipe

Dubbing

Dübverse, Deepdub, DupDub, EZDubs, Papercup, TranslateVideo

Music Editing & Processing

AudioStrip, Cryo Mix, Krisp, LALAL.AI, Moises, VocalRemover

Speech Editing & Processing

Cleanvoice, Podcastle, Resemble.ai Speech-to-Speech, Vocal Clarity, VoiceMod

Coding Applications

Website Generation from Text

Aspen, Debuild, Durable, Enzyme, Galileo AI, Sitekick, Studio AI

Website Generation from Figma Designs

Locofy, Codis, DhiWise

Website Personalization & Optimization

Aidaptive, Intellimize, The.com, uKit AI

Code Generation/Completion

aiXcoder, Bito, CodeAssist, CodeComplete, CodeGPT, CodeI, Codeium, GitHub Copilot, Mutable, Replit Ghostwriter, Tabnine, Warp AI

Code Analysis & DevOps Intelligence

Adrenaline, AI Code Reviewer, Codacy, Codeball, Coderbuds, Codiga, Metabob, What The Diff, Whispr

Documentation Generation

DocumentationLab, DocuWriter, FigStack, Mintlify, Stenography

Data Applications

Automated Analysis & Insights

AI2sql, Channel, Chat2Stats, Consensus, Dataherald, Defog, GenerativeBI, Finalle, Kanaries, Lookup, Maya, ProbeAI, SQL Genie, SQL Genius, String, Symbl, TableTalk, Windsor

Machine Learning & DataOps

Airops, aporia, beam , Deepset, FeatureForm, Gantry, Lang, Jina

Synthetic Training Data Generation

Datagen, gretel, Mirry, Syntegra, Synthesis AI, MostlyAI

Bots

Chatbots

Ada , Air Odin, Balto, Botsify, Certainly, Cohere, Cresta, Elise, Forethought, Haptik, IBM Watson Assistant, Intercom, Kasisto, Observe AI, Sapling, Tymely, Ultimate, Verloop

Personal Assistants & AI Agents

Adept, AgentGPT, Auto-GPT, BabyAGI, Incentivai, Merlin, PolyAI , prodigy AI, shepherd, Stork, Unique AI, Ushur, Xembly

Other Applications

Gaming

Inworld, charisma, Hidden Door narrative AI, latitude, Poly 

Drug Development

JadBio, Lavo, PharmaTrace, Potion

Language Learning

Duolingo, Langotalk, Language Mate, LearnLingo, Lingobo, Praktika, Quazel, Saruna, Slai, Soofy, Tandem GPT

Miscellaneous

AnthemScore, Bladerunner, Hexo, Looria, Mixo, Salesforce Einstein

Continue to Part 2. Investments, Fundraising, Insights and trends.