Antonio Garcia provides a Marxist perspective and programmatic response on AI. This is Part 2 of 2. You can read Part 1 here.

Marxism: Against Reaction, Activism, and Moralism
The Activist Assertions
We now shift our gaze to the tactics employed by the anti-AI left. As I mentioned before, the majority of anti-AI ‘organising’ takes place entirely online. We see individuals utilising the soapbox of social media to voice their concerns and criticisms about the technology. We see groups attempting to uncover users of AI for the purposes of ‘canceling’ such ‘artists.’ We see calls for boycotts against companies that utilise AI in their advertising or products, but once again, these actions only take place behind screens. There are then two layers to our critique of such tactics. First and foremost is the obvious criticism: that substituting real, in-person, organising for online debate is ineffective. Secondly, we must emphasise the weaknesses of activism as a strategy.
What needs to be said regarding the ineffectiveness of online ‘organising?’ It is entirely immaterial. Every callout post, every long-winded rant, is just words on a screen. On its own, it does not change anything, and the vast, vast majority of those writing such statements do not follow up their words with any tangible action. There is little qualitative difference between writing posts and screaming into a field. Cancelling and boycotting users of AI, similarly, achieves little. If the goal is to stop consumption, these activists contradict themselves before they even ‘act.’ Social media platforms make their money from user engagement, so rallying people to not consume through such services is inherently self-defeating. This is particularly poignant as many of AI’s most enthusiastic adopters are social media platforms. From X’s Grok to Meta AI, the very act of using social media financially supports the use of AI. Attempting to rally boycotts of the technology through such platforms are thus a self-defeating tactic.
Now, regarding the matter of activism, which will doubtlessly be of greater controversy. Activism represents a movement’s surrender to idealism. It is, to quote Bordiga’s excellent article, “an illness of the workers’ movement which requires continuous treatment.” Activism fundamentally misunderstands the relations of force in society, believing that the words of the disgruntled masses stand toe-to-toe with the arms, institutions, and economic weapons of the bourgeois state. It promotes a purely mechanistic worldview, one wherein a certain quantity of people being upset or ‘made aware’ necessarily leads to political success. It overestimates the power of spontaneous organising, unable to see the difference between the sudden tremors of the roused masses and the precise, well-planned activity of the organised working class. In this regard, the anti-AI left is doubly impotent. It is first incapacitated by its activistic approach, which misconstrues foundational elements of political struggle, then by its application of this incorrect theory solely to the online sphere.
Activism is a testament to the total conquest of bourgeois ideology over the workers’ movement. It indicates the substitution of serious, revolutionary theory for that of opportunists. It is the triumph of the myth that the state is a ‘neutral, mediating,’ force, and is capable of being reshaped by simply announcing the people’s concerns to it. It is thus the rejection of the state as a tool of its ruling class in favour of the bourgeois narrative- and in doing so, a total rejection of Marxism.
Furthermore, the mechanistic worldview which activism promotes is steeped in the political processes of the bourgeois state. The masses are reminded at every turn that our present society is a mechanistic one. The entire system is reduced to numbers; elections are won by 50.1% of people choosing one candidate over another, petitions succeed when 50.1% (and the bourgeois state) agree to enact something, bills only pass parliament when a certain number of MPs consent, and their right to vote is only granted when a certain number of them choose to show up. It is then obvious why activists utilise these same methods in their struggles: these are the only tactics they know. They then incorrectly apply the courtly proceedings of bourgeois governance to the entire sphere of politics, appalled when such procedures do not spark revolutionary change. To be sure, the workers’ state will doubtlessly contain similar institutions which may appear, and may at first be, similarly mechanistic. However, this is not the topic of the article. Whilst there is much to critique in the organisation of democracy, such criticisms must be spared for another time.
In a similar manner to mechanism, we find that activism promotes a misunderstanding of another key concept in organising: spontaneity. Rather than promoting the bourgeois conception of the state, a propagandised view of reality, this misunderstanding stems from general political illiteracy. Activism, overall, is a reactionary movement. I do not say this to mean they are politically-economically backwards, though the petite-bourgeois aspirations of many such movements often makes them so, but rather, activism relies on reacting to crises of bourgeois society. For example, Black Lives Matter protests have not historically arisen out of a carefully plotted, well-calculated agenda set by a centralised organisation, but in response to acts of police brutality. Activism surrounding climate change did not rise until the late twentieth century, despite the science behind the phenomena existing since the 1950s. Moreover, much of its early activism centred on stopping the construction of mines, logging operations, and other localised phenomena. It was only around the turn of the millennium that activism surrounding global climate change emerged, and it is obvious that this movement is not centralised, nor particularly well-coordinated.
Anti-AI activism follows a similar course- activists have consistently remained several steps behind the development of the technology. It was only after AI was widely adopted that these activists began their work, and by then it was too late. Activism thus promotes a reactionary, spontaneous approach. It falsely believes that the haphazard, sudden wrath of the masses is enough to overturn capitalism- or at least, the elements of capitalism said activists don’t like. This belief is as silly as expecting a sleepwalker to run a marathon; whilst their uncoordinated movements may take them somewhere, they won’t arrive at the finish line without a conscious mind tracking the path ahead.
We thus find that the activist tactics of the anti-AI left are rooted in both bourgeois ideology and political immaturity. It, on the one hand, imitates many of the bourgeois state’s methods, improperly applying them to the workers’ movement. Moreover, activism believes the propaganda the state spins about itself, and thus seeks to utilise its ‘neutral’ stance and mechanistic democracy for ‘revolutionary’ ends. This stifles any earnest revolutionary theory and practice within the workers’ movement, leaving only opportunists and their ideologies in its wake. On the other, activism promotes tactics that simply don’t work. It believes that angry Instagram reels and Reddit posts are praxis, and it lauds a spontaneous, reactionary, approach to politics that has failed time and time again. At such a terminal stage, whether the movement is run by authentic revolutionaries or opportunists is irrelevant, for its tactics are so backwards that it will not accomplish anything regardless.
The Idealist Assertions
The final leftist critique we shall examine stems from idealistic arguments. There are, of course, viewpoints that appeal to one’s righteous anger at the job loss occurring among the proletariat and petite-bourgeoisie. On this front, I have already put forth a materialist account of the phenomena, and the critiques I will soon lay out will cover the idealistic flank of this point. The remaining moralistic arguments demonstrate the correctness of Marx’s analysis of idealism. Of the anti-AI arguments, we shall focus on two: that AI is evil because it ‘steals’ artists’ artwork, and secondly, how AI relates to ‘the spirit of art.’ These arguments represent the trajectory of ideals from production to principles, thus serving as an illuminating modern case study of Marx’s theories.
In his critique of moral philosopher James Mill, Marx laid out his theory of how individualism as an ideal emerged from capitalist production. He described the process in five steps, as listed below:
- Wage labour necessarily alienates workers from their labour. They are disconnected from their labour through their lack of control over production, which takes place as an exploitative extraction of said workers’ surplus value. They are thus detached from their daily activities.
- This alienates workers from the products of their labour. They are thus detached from the products of their daily activities.
- This relationship takes place in a context where workers are impoverished, forced to consent to their exploitation by their need for money and the necessities it can purchase. They are thereby enslaved by their needs.
- The perpetuation of this alienated existence thus becomes the end goal of the workers’ activity. They work so they may survive.
- Alienated from their lifestyle and forced to focus solely on their own needs, workers come to exalt the individual and ego. This self-centredness then manifests politically and philosophically as individualism.
In short, Marx’s theory was that individualism, as well as other prevailing principles of our day, stem from production and the relationships therein.
We find a similar process of idealisation emerging from production today. This principle presents itself as an opposition to the ‘theft’ AI commits through its use of existing artwork as a base for its outputs. I shall lay out the progress of this principle in the same manner Marx did:
- Art under capitalism is commodified as intellectual property. Ideas and creations are the capital of those who create them.
- This forces most professional artists into the petite-bourgeoisie.1 In order to make any sort of money, they must capitalise on their art, and capitalisation is necessary if they want to survive.
- Intellectual property serves as the basis for artists’ ability to turn a profit. They therefore become highly defensive of such regulations, as doing so is the only way to protect their lifestyle and capital.
- Artificial intelligence contradicts intellectual property law by blurring the lines between inspiration and theft. It does this by undermining artists’ ability to sell their art, as it makes similar works free and easily reproducible. It also undermines its status as capital, as allowing art to be freely redistributed and/or recreated lowers its exchange value.
- Artists thus press the claim that AI is ‘stealing’ their art because it imitates their capital. This is both a necessity to conserve their capital, and gives way to outrage at technology’s ‘audacity’ to ‘steal’ from them.
Thus, the interests of professional artists are bound to the perpetuation of intellectual property- an outgrowth of capitalism. They therefore must oppose all attempts to discredit or abolish this form of capital, be it through piracy, replication, or AI’s use of their art. Hence why nominally progressive artists often oppose these acts. Whether from personal experience or from that of others, artists are ingrained with the idea that any imitation or unprofitable transmission of their work is an affront to themselves. The matter then transcends a mere matter of business and becomes a personal affair. Such principles explain, justify, and perpetuate professional artists’ petite-bourgeois privilege. They arise from the processes of capitalist production, conveniently serving their interests, thus succeeding in establishing an idealised framework for what art is and how its market ought to function.
We then see the claim that AI ‘art’ diminishes the ‘spirit of art’ to be a hasty attempt to integrate new ideals into this framework. Simply put, what is ‘the spirit of art?’ No one knows. The claim of ‘soullessness’ is levied against every art form by its detractors. Abstract and Expressionist works were deemed ‘meaningless’ and ‘degenerate’ by their critics as styles such as Corporate Memphis are often directly called ‘soulless’ today. Currently, the ‘AI art style,’ inasmuch as one exists, currently faces similar critiques. AI ‘art’ is criticised for being illogical and visually unappealing, for being “emotionally empty,” or being too easy to produce. On the other side, those in favour of AI ‘art’ herald its greatness for helping people “[infuse] technology with their inner drive,” or that it can lead to one transcending “the limitations of human perception and cognition.”
These claims vary from subjective to nonsensical. The arguments of those against AI ‘art’ are simply criticising stylistic decisions they don’t like. Plenty of art is illogical and visually unappealing, intentionally or not. Plenty of art lacks emotional depth. Plenty of art doesn’t require much skill to produce. These criticisms are exactly the sort levied at art-forms such as abstractionism, which is often denounced along similar lines. We thus find that the AI element of such ‘art’ is, in reality, rarely the topic of discussion. Rather, these detractors are criticising AI outputs in much the same way other art-forms have been examined. In this sense, such advocates contradict themselves: they claim that AI outputs aren’t art whilst treating them like any other artwork. Instead of criticising AI outputs though separate, non-artistic, forms of criticism, they use the terms and ideas of the art world to condemn them, thereby unintentionally elevating AI outputs to the status of art.2
On the other hand, what does ‘infusing technology with one’s inner drive’ mean? That the AI is possessed by the user’s soul when it generates an output? Any claim by the pro or anti-AI ‘art’ crowd that art is imbued with ‘spirit,’ ‘soul,’ or ‘heart,’ necessarily appeals to spiritual ideals about human existence. The belief that one’s spiritual essence somehow characterises or possesses anything is absurd. The soul lacks any material evidence to prove its existence; thus, to claim art is possessed by the soul is no different to claiming it is possessed by faeries, demons, or ghosts. And the claim that AI outputs “transcend the limitations of human perception and cognition-” does the author, presuming a human wrote this, understand what these words mean? If an output genuinely escapes perception and cognition, we couldn’t recognise it. We would have no idea of its existence. To transcend perception is to be unable to hear, see, touch, or otherwise sense something. To transcend cognition is to be, quite literally, incomprehensible. Artificial intelligence is incapable of doing this right now, and if it ever became capable of such, we wouldn’t be able to know. Such a claim is not only demonstrably false- most people can perceive AI outputs quite well -it is also gibberish. It loses all sense if one spends even a moment considering it.
As one can see, these attempts to criticise or commend AI ‘art’ are firmly rooted in idealism. Part of this is a necessary consequence of art- one’s taste in ‘good’ and ‘bad’ art is determined by innumerable factors, from their personal taste to the reason they’re seeking art in the first place. It is a subjective matter with little science behind it. I believe another part of this reliance on idealism stems from the emergent nature of this debate. AI, as it currently stands, did not exist a decade ago. Even now, from the start of the AI craze in 2023 to the present, we have seen rapid development in the technology. As AI continues to develop at a break-neck pace, it is difficult to ‘nail it down’ and carefully examine. Thus, whilst material changes in production become rapidly apparent, leading to the outspoken backlash of at-risk workers and capitalists, it has not yet had the time to fully settle into our cultural-intellectual zeitgeist. Until its integration is complete, the best the bourgeoisie can do is regurgitate rehashed ideals until something sticks. Hence, whether one is dealing with pro or anti-AI advocates, they are bound to be swamped by idealist assertions. As Marxists, we cannot accept such simplified analyses. We must instead grapple with reality as it is, a process which I hope to have started, so we may construct a response capable of serving the interests of the proletariat and its historic task.
A Brief Note on AI, the Environment, and Marxism
One of the other significant concerns surrounding artificial intelligence is its impact on the environment and natural resources. The amount of water and electricity needed to maintain AI systems is enormous, and as the technology is further adopted and developed, this consumption can only grow. As it currently stands, even a ‘small’ one-megawatt data centre consumes 26 million litres of water each year, and though data centres officially only make up a minute amount of the US’ annual water consumption, roughly one seven-hundredth, nearly one-third of data centres do not measure how much water they consume, greatly muddying this statistic. In Sydney, this concern has arisen with recent projections that claim data centres will consume upwards of one-quarter of the city’s clean water supply by 2035. As climate change’s effects deepen, the threat of drought and the conflict between tech companies and the local population for water make such a prospect difficult to ignore.
In addition, data centres are projected to exacerbate carbon emissions, and thus, the climate crisis. Currently, data centres account for approximately one percent of carbon emissions; by 2030, this is projected to rise as high as eight percent. The high amount of electricity consumed by data centres also raises ominous questions for local communities. In Texas, the electrical grid was overwhelmed by increased usage during a blizzard, resulting in rolling blackouts across the state. 210 people died during this storm, and almost all recorded deaths were found to be related to the blackouts. In a state where much of the power grid serves data centres, the competition over limited resources once again rears its head. This was not an isolated incident, either. In Virginia, another hub for data centres, a blackout was narrowly avoided when dozens of data centres switched to and from backup power systems, resulting in massive shocks to the electrical grid. In 2024, it was reported that Americans experienced almost twice as many blackouts as in 2023. Whilst most of these came from natural disasters like hurricanes and volcanic eruptions, the strain put on electrical grids by AI data centres cannot be ignored.
What is the Marxist position on such issues? As discussed before, we cannot descend into primitivism. We cannot wholly reject the development of technology, and therefore, the mode of production, because of environmental concerns. Such Luddite views were criticised by Marx and Marxists over a century-and-a-half ago, and the underlying logic still applies today. Advancements in technology under bourgeois society can be used to serve the workers’ interests, and often incidentally do. Whilst industrialisation, for instance, was spurred by capitalists’ desire for ever-growing profits, it also improved the workers’ quality of life, led to breakthroughs in medicine, agriculture, and art, and solidified the proletariat as a class, accelerating the working class’ development and bringing its final victory ever closer. Similarly, AI has been useful in detecting previously undiscovered tumours and lesions in patients, and training is underway to teach it how to detect early stages of cancer, Alzheimer’s, and deadly infections. As well, we find through our previous discussions that AI is set to support the working class within certain sectors of the global economy, and that it is already simplifying production, a vital step toward both socialism and communism. Such developments have utility in both bourgeois and proletarian societies, and thus, cannot be wholly disregarded.
On the other hand, the future socialist state will have no need for many of the frivolities of AI. It will not need to maintain data centres for ChatGPT, DeepSeek, CoPilot, and so on, when each system provides the same service. Under organised, de-commodified production, such competition will serve as nothing but a hindrance, and will be abolished. Socialist society will have no qualms with limiting the amount of power and water AI systems may use, as the conflict between profits and community wellbeing will no longer exist. In a society where the commodity form has been abolished, profit will not exist, and thus, there will be nothing but community wellbeing to look after.3 Thus, through a combination of calculated reductions in water and electricity consumption, as well as a fundamental reorientation of economic policy, the socialist state will be armed with several means to respond to the problems of AI without resorting to primitivist self-sabotage.
Our Programmatic Response
It is for these reasons that a programmatic response is needed for the question of artificial intelligence. The workers’ movement must take a mature approach, neither ignoring the plights of workers displaced by AI nor pitting workers of different fields against each other. This response must not be activistic in nature, as doing so will sabotage our efforts from the start. No, we must instead follow a well-planned and well-coordinated approach. Finally, we must eschew the ideals surrounding AI- but we must understand their roots, too. By comprehending the existing ideals around AI and how they insufficiently attempt to relieve the contradictions of capitalism, we can adequately criticise them. We can reveal their farcical nature to the proletariat and build political maturity among workers. All the same, we must acknowledge that AI is very much still developing. We must thus adopt a programmatic response, one that is resolutely Marxist at all stages and is capable of adapting to new advancements in the technology.
It is for these reasons that I propose the following points to outline what a programmatic response ought to be. Whilst it can, and should, contain more, these are the bare-minimum to avoid opportunism, mismanagement, and idealism.
- As Marxists, we must understand the historic role of artificial intelligence, which is as follows:
- AI is an innovation in the realm of machinery. It will therefore inevitably displace a certain subsection of the proletariat and bourgeoisie as it penetrates the market.
- At the same time, artificial intelligence will also drive demand for the technology and its upkeep. Thus, whilst a section of the workforce will be made irrelevant, those segments which construct, supply, or maintain AI systems will see growth.
- It is apparent that AI simplifies the production process, and may further do so pending future advancements. It therefore serves a historically progressive role. This is because it proletarianises a segment of the bourgeoisie, centralises capital, aids future proletarians in managing the socialist state and economy, and serves to diminish the division of labour, thereby serving as a stepping stone toward communism.
- As Marxists, we must ensure our programme serves the proletariat in its historic task. We then draw the following conclusions:
- We cannot reduce our response to primitivism. Whilst we must express our solidarity with those workers whose jobs are made redundant by the technology, we must use this as a case study to such workers. That capitalism has warped a technology that could improve the lives of workers into a tool of their abuse demonstrates that capitalism cannot serve the interests of the working class, nor does it have any sympathy for those it displaces. The enemy is not technological progress, but the mode of production itself.
- We cannot reduce our response to trade union economism and nationalism. We will not pit the workers who have been displaced by AI against those who receive boons from the technology. This applies both domestically and internationally. As the world is dominated by the imperial stage of capitalism, we acknowledge that the division of labour is globalised. Thus, whilst workers in the finance and service-centred economies of the ‘first world’ may experience displacement, there will be boons in economies predicated on industry. We thus cannot tolerate any division of the working class along different industries within Australia, nor between Australian workers and those of other countries.
- We cannot reduce our response to activism. We must recognise that simply voicing one’s discontent or ‘spreading awareness’ accomplishes little on its own. We must combine competent political education with robust political struggle. This necessarily takes the form of a centralised, well-coordinated workers’ party directing the political struggles of the proletariat, and it must have a single goal in mind: the construction of a socialist state which will wither away into communism.
- We cannot reduce our response to solely online activities. This is the highest form of activism, and thereby demonstrates most acutely its contradictions and failings. Only through well-planned and well-organised real-life activity will political change occur.
- We cannot reduce our response to a partnership with the petite-bourgeoisie. It is this class of the most reactionary capitalists which has historically undermined the proletarian movement more than any other foe. Thus, though the discontents of AI may make it appear as though the workers and small proprietors share common interests, we must reject any alliance with them. To align with them is to willingly subjugate the workers’ movement to the interests of the most backwards segment of the bourgeoisie. If individual petite-bourgeois wish for our solidarity, they can betray their class and, after demonstrating their loyalty to the proletariat, willingly subject themselves to the proletarian programme and the responsibilities therefrom.
- We must acknowledge that these failures are not solely the fault of external actors. That opportunist, reactionary, and incompetent dogmas have festered in the proletarian movement is evidence that the workers’ leadership have failed to keep such threats at bay. We must therefore ensure that the workers’ party provides political education of the highest excellence, and that it carefully vets potential cadres prior to assigning them any important tasks.
- As Marxists and materialists, we condemn the idealist approaches to artificial intelligence. We will not baulk at the technology for ‘destroying the spirit of art,’ nor will we imbue it with fantastic assertions better suited for a sci-fi movie than a party programme. Our priority is serving the workers in their historic task. Thus, we must approach the matter from a strictly materialist standpoint. We will express solidarity and deliver aid to workers displaced by AI. We will educate the proletariat about the historic role of machinery. We will build a party capable of organising the workers beyond the scope of trade union economism or activist chicanery.
- We must acknowledge AI’s impact on the environment within a Marxist framework. The science is obvious: the climate is changing, and AI’s high consumption of water and electricity is doubtlessly harmful for the environment. At the same time, the current environmental movement is firmly bourgeois in terms of tactics and demands. It is therefore imperative to construct socialist response to the climate crisis that does not fall into green liberalism, activism, or primitivism.4
- We must acknowledge the still-developing character of AI. Whilst it is currently a mere machine, one that simplifies labour but accomplishes little more, this may not always be the case. In the future, we may find it develop further, displacing more workers and reorganising world economy to a greater extent. Or, we may find the inverse- a failure of the technology to meaningfully alter capitalism. Perhaps even other alternatives will take place- we do not know. Practically, this means we must maintain a programmatic doctrine toward AI. Whilst the above points are non-negotiable for a Marxist approach to the issue, the strategies of the workers’ movement must be, at all stages, be attuned to capitalism as it stands. Thus, we must keep an open mind to strategies both old and new.
With these points in mind, I hope to have created a suitable outline which you, the readers of Partisan! may build upon to construct a capable approach to artificial intelligence and its future. I wish to personally express my enthusiasm for future discussions surrounding Marxism, its tactics, and its approach to modern technologies, regardless of whether these points are integrated into any statement or programme. So long as the discussion is cordial, well-informed, and rooted in Marxist theory and practice, I cannot ask for more. I wish you, the readership, the best as we continue our relentless pursuit of a brighter world.
Footnotes
1 To be clear about my terminology, I use ‘professional artists’ to mean individuals who sell their art to make a living, not hobbyists or the like. I also use the word ‘most,’ as, whilst most professional artists are self-employed or are otherwise commissioned for their work, there are some who operate as proletarians, such as those hired by companies to create art on a regular basis, such as game designers, architects, and animators at certain companies. This said, the ideals surrounding art penetrate all classes and professions within the community. Thus, it is entirely possible for a hobbyist or proletarian artist to echo these sentiments even though they themselves are not bourgeois. As Marx and Engels wrote, “the ideas of the ruling class are in every epoch the ruling ideas.”
2 To further clarify this point, perhaps an example would be of use. Consider two scenarios: appreciating a painting at an art gallery, and waiting in line to get a driver’s license. Both the painting and the line can be criticised, but the forms of critique are obviously not the same. Criticising a painting as being ‘too long’ makes no sense, and criticising waiting in a queue as ‘requiring no skill’ is similarly absurd. This, of course, doesn’t mean there is absolutely no overlap; one can call both a painting and a bureaucratic process ‘soulless’ and mean similar things. However, these fleeting similarities do not make waiting in line for a driver’s license ‘art’ by any reasonable understanding of the word.
It is on this point where critics of AI outputs fail. In criticising AI outputs with the same terms and lenses as art, they implicitly equate the two. Whilst they rail and rave about how AI outputs are not art, they treat them like art by virtue of how they critique them. Such arguments are intrinsically contradictory, and the only way to straighten them out is to categorise AI outputs as art, which is antithetical to their point, or to disregard all elements of their argument except the conclusion, which is also unsatisfactory. Thus, we find that such arguments are unacceptable not only from a Marxist-materialist perspective, but also from a simpler breakdown of the claims being made.
3 For a deeper discussion of socialist economics and the imperative for the abolition of commodity production, I suggest my article in the Prometheus Magazine, The Question of Profit, as well as Bordiga’s Dialogue with Stalin, Bukharin’s The New Economic Policy of Soviet Russia, and, though perhaps a bit too broad, the first volume of Marx’s Capital.
4 To this end, I believe the analyses put forth by Max J. hold many of the same positions vis-a-vis the environmentalist movement I wish to encourage. As recent experience has demonstrated, environmental activism stands at-odds with the organisation and tactics of the workers’ movement. Moreover, he understands that productive forces will necessarily advance under socialism and communism, and that this is not an “unfortunate temporary measure” but a necessary outcome of any revolution in production. We thus find, albeit through different reasoning and language, that the workers’ movement cannot rely on the tactics of activists, nor is it theoretically compatible with the opportunist and primitivist views environmentalists currently hold.




You must be logged in to post a comment.